Available via license: CC BY
Content may be subject to copyright.
RESEARCH ARTICLE
National and regional prevalence of
posttraumatic stress disorder in sub-Saharan
Africa: A systematic review and meta-analysis
Lauren C. NgID
1
*, Anne StevensonID
2
, Sreeja S. Kalapurakkel
3,4
, Charlotte HanlonID
5
,
Soraya Seedat
6
, Boniface HarerimanaID
7
, Bonginkosi Chiliza
8
, Karestan C. KoenenID
9
1Department of Psychology, University of California Los Angeles, Los Angeles, California, United Statesof
America, 2Department of Epidemiology, Harvard T.H. Chan School of Public Health, Boston,
Massachusetts, United States of America, 3Duke University Global Health Institute, Durham, North Carolina,
United States of America, 4Centre for Global Mental Health, Health Service and Population Research,
Department Institute of Psychiatry, Psychology and Neuroscience King’s College, London, United Kingdom,
5Department of Psychiatry, Stellenbosch University, Cape Town, South Africa, 6Faculty of Health
Sciences, Western University, London, Ontario, Canada, 7College of Medicine and Health Sciences,
University of Rwanda, Kigali, Rwanda, 8Department of Psychiatry, Nelson R. Mandela School of Clinical
Medicine, University of KwaZulu-Natal, Durban, South Africa, 9Department of Epidemiology, Harvard T.H.
Chan School of Public Health, Boston, Massachusetts, United States of America
*laurenng@ucla.edu
Abstract
Background
People living in sub-Saharan Africa (SSA) are disproportionately exposed to trauma and
may be at increased risk for posttraumatic stress disorder (PTSD). However, a dearth of
population-level representative data from SSA is a barrier to assessing PTSD. This manu-
script sought to calculate pooled PTSD prevalence estimates from nationally and regionally
representative surveys in SSA.
Methods and findings
The search was conducted in PubMed, Embase, PsycINFO, and PTSDpubs and was last
run between October 18, 2019, and November 11, 2019. We included studies that were
published in peer-reviewed journals; used probabilistic sampling methods and systematic
PTSD assessments; and included 450 participants who were current residents of an SSA
country, at least 50% of whom were aged between 15 and 65 years. The primary outcomes
were point prevalence estimates of PTSD across all studies, and then within subgroups.
The protocol was registered with the International Prospective Register of Systematic
Reviews (PROSPERO) (registration number CRD42016029441). Out of 2,825 unique arti-
cles reviewed, 25 studies including a total of 58,887 eligible participants (54% female) in 10
out of the 48 countries in SSA were identified. Most studies enrolled any adult aged 18 years
or older. However, some studies only enrolled specific age brackets or persons as young as
15 years old. Six studies were national surveys, and 19 were regional.
There were 4 key findings in the meta-analysis: (1) the overall pooled prevalence of prob-
able PTSD was 22% (95% CI 13%–32%), while the current prevalence—defined as 1 week
PLOS MEDICINE
PLOS Medicine | https://doi.org/10.1371/journal.pmed.1003090 May 15, 2020 1 / 30
a1111111111
a1111111111
a1111111111
a1111111111
a1111111111
OPEN ACCESS
Citation: Ng LC, Stevenson A, Kalapurakkel SS,
Hanlon C, Seedat S, Harerimana B, et al. (2020)
National and regional prevalence of posttraumatic
stress disorder in sub-Saharan Africa: A systematic
review and meta-analysis. PLoS Med 17(5):
e1003090. https://doi.org/10.1371/journal.
pmed.1003090
Academic Editor: Peter Byass, UmeåCentre for
Global Health Research, UmeåUniversity,
SWEDEN
Received: August 2, 2019
Accepted: April 13, 2020
Published: May 15, 2020
Copyright: ©2020 Ng et al. This is an open access
article distributed under the terms of the Creative
Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in
any medium, provided the original author and
source are credited.
Data Availability Statement: Due to the nature of
the meta-analysis, all data are present in the paper
and are therefore publicly available.
Funding: This study was supported by the National
Institute of Mental Health (L.N., #K23MH110601),
by Cohen Veterans Bioscience (K.C.K), by the UK
Department for International Development (DfID)
(C.H., #201446) as part of the Programme for
Improving Mental health carE (PRIME), by the
to 1 month—was 25% (95% CI 16%–36%); (2) prevalence estimates were highly variable,
ranging from 0% (95% CI 0%–0%) to 74% (95% CI 72%–76%); (3) conflict-unexposed
regions had a pooled prevalence of probable PTSD of 8% (95% CI 3%–15%), while conflict-
exposed regions had a pooled prevalence of probable PTSD of 30% (95% CI 21%–40%; p
<0.001); and (4) there was no significant difference in the pooled prevalence of PTSD for
men and women. The primary limitations of our methodology are our exclusion of the follow-
ing study types: those published in languages other than English, French, and Portuguese;
smaller studies; those that focused on key populations; those that reported only on continu-
ous measures of PTSD symptoms; and unpublished or non–peer-reviewed studies.
Conclusions
In this study, PTSD symptoms consistent with a probable diagnosis were found to be com-
mon in SSA, especially in regions exposed to armed conflict. However, these studies only
represent data from 10 of the 48 SSA countries, and only 6 studies provided national-level
data. Given the enormous heterogeneity expected across the continent, and also within
countries and regions, this review cannot speak to rates of PTSD in any regions not included
in this review. Thus, substantial gaps in our knowledge of PTSD prevalence in SSA remain.
More research on population-level prevalence is needed to determine the burden of trauma
symptoms and PTSD in SSA and to identify acceptable and feasible approaches to address
this burden given limited mental healthcare resources.
Author summary
Why was this study done?
• Repeated and prolonged exposure to violence, armed conflict, and mass-casualty events,
combined with a lack of access to mental health treatment, may result in a substantial
effect on the population burden of posttraumatic stress disorder (PTSD) in sub-Saharan
Africa (SSA).
• While many studies of PTSD have been conducted in SSA, most of these studies derived
their estimates from nonrepresentative samples or specific populations.
• Population-representative epidemiologic data are critical to understand the burden of
PTSD in SSA and develop national and regional policies to address that burden.
What did the researchers do and find?
• We conducted a systematic review and meta-analysis of the prevalence of PTSD from
representative national and regional studies in SSA.
• Pooled prevalence estimates were calculated across all studies, and then within sub-
groups including by sex, assessment time frame (i.e., 1 week, 1 month, 1 year), use of a
screening or diagnostic measure, and whether populations were affected or not affected
by mass-casualty war or armed conflict.
PLOS MEDICINE
Meta-analysis of prevalence of PTSD in sub-Saharan Africa
PLOS Medicine | https://doi.org/10.1371/journal.pmed.1003090 May 15, 2020 2 / 30
National Institute of Health Research (NIHR) Global
Health Research Unit on Health System
Strengthening in Sub-Saharan Africa, King’s
College London (C.H., GHRU 16/136/54), and by
AMARI as part of the DELTAS Africa Initiative (C.H.,
DEL-15-01). The views expressed in this article do
not necessarily reflect the UK Government’s official
policies, those of the NHS, the NIHR, or the
Department of Health and Social Care. The funders
had no role in study design, data collection and
analysis, decision to publish, or preparation of the
manuscript.
Competing interests: C.H. is a member of the
Editorial Board of PLOS Medicine.
Abbreviations: CIDI, Composite International
Diagnostic Interview; DRC, Democratic Republic of
Congo; DSM-IV, Diagnostic and Statistical Manual
version 4; DTS, Davidson Trauma Scale; EVD,
Ebola virus disease; HTQ, Harvard Trauma
Questionnaire; HTQ-R, Harvard Trauma
Questionnaire-Revised; ICTR, United Nations’
International Criminal Tribunal for Rwanda; IDP,
internally displaced person; IES-6, Impact of Event
Scale-6; MeSH, medical subject heading; mhGAP,
Mental Health Gap Action Programme; MINI, Mini
International Neuropsychiatric Interview; PCL-C,
PTSD Checklist-Civilian Version; PDS,
Posttraumatic Stress Diagnostic Scale; PRIME,
Programme for Improving Mental Health Care;
PRISMA, Preferred Reporting Items for Systematic
Reviews and Meta-Analyses; PROSPERO,
International Prospective Register of Systematic
Reviews; PSS-I, PTSD Symptom Scale Interview;
PTSD, posttraumatic stress disorder; SCID,
Structured Clinical Interview for DSM Disorders,
Axis I, Research Version; SSA, sub-Saharan Africa;
TSQ, Trauma Screening Questionnaire; WMH,
World Mental Health.
• We identified 25 unique studies (N= 58,887) across 10 of the 48 SSA countries.
• The pooled prevalence of probable PTSD across all studies was 22% (95% CI 13%–32%).
• The pooled prevalence of probable PTSD in participants from war-exposed regions was
30% (95% CI 20%–40%), while the estimate from war-unexposed regions was 8% (95%
CI 3%–15%; p= 0.01).
What do these findings mean?
• These data suggest that PTSD symptoms and probable PTSD are common in SSA.
• However, information was only found on 10 of the 48 SSA countries, and only 6 studies
provided national-level data.
• Only one study used a measure of PTSD symptoms whose reliability and validity had
been assessed previously in the population of interest.
• Our results suggest both that PTSD is a major public health problem in SSA and that
large gaps in our knowledge of this problem remain.
Introduction
Mental and substance use disorders account for 23% of years lost to disability, making them
the leading cause of disability worldwide [1]. Posttraumatic stress disorder (PTSD) is a large
contributor to the global burden of disease and is estimated to affect almost 4% of the world’s
population [2]. PTSD persists for over a year in 50% of all cases [2] and often leads to substan-
tial declines in functioning and productivity [3]. National and regional data on prevalence are
used to develop policies and action plans for addressing PTSD and other related disorders.
Recently, much of the data on global and national estimates of PTSD has come from the
World Health Organization’s World Mental Health (WMH) surveys of the cross-national
prevalence of PTSD in 26 countries [2,4,5]. The WMH surveys collected representative popu-
lation data across the world using structured diagnostic measures to assess PTSD, which
allowed for the calculation of global population prevalence estimates of PTSD [2]. However,
the WMH surveys only included one national survey and one regional survey from sub-Saha-
ran Africa (SSA); in addition, the one national estimate was from South Africa, one of the few
upper-middle–income countries in SSA [2]. The limited number of countries from SSA con-
tributing data on PTSD in the WMH surveys is consistent with the lack of population mental
health data from SSA generally [6,7].
People living in SSA may be disproportionately affected by individual and population-level
trauma exposure. Indeed, research from the World Health Organization has found that the
lifetime prevalence of road traffic deaths [8] and of reported intimate partner violence and/or
nonpartner sexual violence are highest in the Africa region [9,10]. In addition, although other
regions of the world experience more natural disasters, the great majority of countries most
vulnerable to natural disasters are in SSA [11]. SSA has also been disproportionately affected
by war and armed conflicts, many of which have been ongoing for years, if not decades [12].
In 2019, 20 countries in SSA were classified by the World Bank as hosting fragile and conflict-
PLOS MEDICINE
Meta-analysis of prevalence of PTSD in sub-Saharan Africa
PLOS Medicine | https://doi.org/10.1371/journal.pmed.1003090 May 15, 2020 3 / 30
affected situations, which represents more than 50% of the fragile and conflict-affected coun-
tries globally [13]. Moreover, the legacy of violence, loss, and historical trauma inflicted on
the people in SSA through colonization may contribute to high rates of posttraumatic stress
[14].
Increased exposure to traumatic life experiences in SSA is compounded by very low rates of
access to mental health treatment [15,16]. Of the 48 countries in SSA [17], 24 (50%) are low-
income countries, 18 (37.5%) are lower-middle–income countries, and the remaining 6
(12.5%) are upper-middle–income countries [18]. It is estimated that the gap between those
who need mental healthcare and those who receive it often exceeds 90% in low-income coun-
tries for most mental disorders [19–21]. It is estimated that 77% of people with PTSD in lower-
middle–income countries have not received treatment [2]. Given that the majority of countries
in Africa are low-income or lower-middle–income, it is likely that the vast majority of people
with PTSD in SSA will never receive treatment and are at high risk for chronic symptoms.
Repeated and prolonged exposure to violence, armed conflict, and mass-casualty events, com-
bined with a lack of access to mental health treatment [15,16], may result in a substantially
larger effect on the population burden of PTSD in SSA [22,23].
Aims
The goal of this meta-analysis was to synthesize the existing data on the population prevalence
of PTSD in SSA. While many studies of PTSD have been conducted in SSA, most of these stud-
ies derived their estimates from nonrepresentative samples or specific populations such as ref-
ugees or internally displaced persons (IDPs) [24–30], patients [31–35], parents [36–40], or
students [41–49]. While studies have summarized PTSD prevalence in conflict-affected popu-
lations, including those in SSA [50,51], PTSD occurs in response to a wide range of traumas
(both interpersonal and non-interpersonal) that occur in non-conflict settings (e.g., car acci-
dents, sexual assault). Population-representative epidemiologic data are critical to understand
the burden of PTSD in SSA and develop national and regional policies to address that burden.
To our knowledge, this is the first meta-analysis to summarize the data on population-based
point prevalence of PTSD in SSA across all settings. The objectives of this paper were to (a)
conduct a systematic review and meta-analysis of the prevalence of PTSD from representative
national or regional studies and (b) explore the association between sex, population-level expo-
sure to armed conflict, and pooled prevalence of PTSD.
Methods
Search strategy and selection criteria
Study inclusion criteria are as follows:
1. Participants were current residents of a country in SSA (e.g., citizens, permanent residents,
IDPs, or refugees or immigrants who had resided in the country for at least 6 months).
2. In order to capture studies that focused primarily on adults, at least 50% of participants had
to be between 15 and 65 years of age, which is defined as the “working age population” by
the World Bank [52].
3. PTSD data were reported for at least 450 participants. We selected a priori a sample size cut-
off to ensure that we would have enough power to detect prevalence estimates lower than
3.9%, which is the mean lifetime PTSD prevalence across the WMH surveys [2]. A mini-
mum sample size requirement of 450 would allow us to identify prevalence as low as 3.3%
with a precision of 1.65% within each study or region [53].
PLOS MEDICINE
Meta-analysis of prevalence of PTSD in sub-Saharan Africa
PLOS Medicine | https://doi.org/10.1371/journal.pmed.1003090 May 15, 2020 4 / 30
4. Studies used a systematic method of classifying participants as those with or without PTSD.
Studies could be included if they used structured or semi-structured interviews—either
administered by trained lay interviewers or clinicians—to provide a diagnosis or if they
used a symptom checklist or diagnostic assessment and applied a cut-point to classify peo-
ple as having PTSD or not having PTSD. In addition, we included studies that assessed cul-
tural idioms of distress that were clearly related to symptoms developed in response to a
stressful or traumatic event [54]. These are referred to as “post trauma reaction syndromes.”
For the purposes of this paper, any participants who were either diagnosed with PTSD fol-
lowing an interview or who scored above a cut-point criterion set by the study authors were
considered to have symptoms consistent with a diagnosis of PTSD.
5. Studies employed probabilistic procedures to obtain nationally or regionally representative
samples. Representative samples that were limited to a specific sex were also included.
6. Articles were written in English, French, or Portuguese, which covers most of the scientific
literature from SSA.
7. Studies were published in a peer-reviewed journal with no journal date restrictions.
8. If studies included a mixture of eligible and ineligible participants and data from the group
meeting inclusion criteria could be disaggregated, the data that met inclusion criteria were
included in the review.
The search strategy was developed and initially conducted in PubMed between February
22, 2016, and August 1, 2017. The search was expanded to 4 databases and rerun using
PubMed, Embase, PsycINFO, and PTSDpubs between October 18, 2019, and November 11,
2019. Search terms were “[Name of SSA Country] AND PTSD (in any field).” The search was
run for each country in SSA. Using Kenya as an example, PubMed automatically generates the
following search when these terms are entered: (“Kenya” [medical subject heading (MeSH)
Terms] OR “Kenya” [All Fields]) AND (“stress disorders, post-traumatic” [MeSH Terms] OR
(“stress” [All Fields] AND “disorders” [All Fields] AND “post-traumatic” [All Fields]) OR
“post-traumatic stress disorders” [All Fields] OR “ptsd” [All Fields]). The full texts of articles
that appeared to meet inclusion criteria based on the abstract were downloaded, and references
were reviewed to identify additional papers that might meet the eligibility criteria.
The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) [55]
(see Fig 1 for complete PRISMA flow diagram) process was followed to identify and include/
exclude the papers in this review. Citations, abstracts, and full-text articles (when available) for
all potentially eligible articles were downloaded and double coded by a team of 8 researchers to
ensure they met criteria.
Articles were identified through the PubMed, Embase, PsycINFO, and PTSDpubs search
and by reviewing references from the identified papers. Seven fields were then extracted from
each paper and entered into a Google Docs spreadsheet: article title, year, first author, journal,
search date, researcher conducting the search, and abstract. Duplicate entries were flagged and
filtered out of the spreadsheet. To identify eligible studies from the remaining articles, the full
text of each article was reviewed, and one coder entered information related to each inclusion
and exclusion criteria into the database. A second coder independently reviewed each of the
inclusion and exclusion criteria to assess concordance. Disagreements about study eligibility
were reviewed and discussed by the research team, and final decisions were made by
consensus.
The abstracts of non-English articles were reviewed by native speakers of the article lan-
guage to determine eligibility, and if the abstract was deemed eligible, the full article was
PLOS MEDICINE
Meta-analysis of prevalence of PTSD in sub-Saharan Africa
PLOS Medicine | https://doi.org/10.1371/journal.pmed.1003090 May 15, 2020 5 / 30
reviewed by the native speaker and also underwent targeted translation into English to enable
accurate data extractions by a second coder. Studies that were not related to the eligibility crite-
ria (e.g., a commentary or a study describing physical trauma instead of psychological trauma)
were flagged and removed. For studies for which it was unclear whether they met eligibility cri-
teria, the first author (L.C.N.) emailed the authors of the study to clarify.
Study quality was assessed using a standardized tool for observational studies [56]. Loney
and colleagues (1998) was selected against a range of checklists and scales based on its applica-
bility to prevalence studies, validation, reliability, and clear methodology for rating studies
[57]. The tool contains 8 items, each with a possible score of 1, with a maximum total score of
8.
In implementing the quality assessment tool, we operationalized 4 of the questions on the
Loney tool to best suit the purposes of this systematic review and meta-analysis:
1. On the first question (“Are the study design and sampling method appropriate for the
research question?”), a full point was given if the study was a whole population or a random
sample and the authors described the process in enough detail to support the statement
from region down to the randomization of the individual selected to participate. Studies
Fig 1. Study selection flow diagram. PTSD, posttraumatic stress disorder.
https://doi.org/10.1371/journal.pmed.1003090.g001
PLOS MEDICINE
Meta-analysis of prevalence of PTSD in sub-Saharan Africa
PLOS Medicine | https://doi.org/10.1371/journal.pmed.1003090 May 15, 2020 6 / 30
received a 0 if they were not the whole population or a random sample, or a half point if the
study design was described as a “random sample”, but there was not enough evidence to
confirm this.
2. On the fourth question (“Are objective, suitable, and standard criteria used for measure-
ment of the health outcome?”), we defined this as “PTSD measurement tools which are con-
sistent with accepted clinical/research criteria and validated in the population of interest.”
If the battery was an acceptable clinical/research tool but had not been validated during the
study or previously in the population of interest, it received a half point.
3. On the fifth question (“Is the health outcome measured in an unbiased fashion?”), we
adapted this to make it more appropriate to our study framework. The original tool recom-
mends blinding interviewers to the purpose of the study in some cases. We did not feel this
was applicable in the context of prevalence studies of PTSD in SSA as most teams employed
interviewers to conduct the batteries and trained them specifically on research methods in
order to implement the study. Papers received a full point on this question if they described
the assessors and described the training that they received to conduct the interviews and if
internal reliability was calculated. If only one of these items was stated in the paper, the
question was scored as a half point.
4. On the sixth question (“Is the response rate adequate? Are the refusers described?”), papers
received a full point if the response rate was at least 70% and if the authors described any
demographic details about the refusers. Studies scored a 0 if they did not meet the response
rate threshold and did not describe the refusers, or a half point if they did one or the other.
All other items on the Loney tool were scored as either 0 or 1 with no half points. The asso-
ciation between study quality and prevalence estimate was assessed using a meta-regression
analysis.
This systematic review was registered in the International Prospective Register of System-
atic Reviews (PROSPERO) on December 1, 2016, and updated on June 21, 2019 (registration
number CRD42016029441; https://www.crd.york.ac.uk/prospero/display_record.php?
RecordID=29441&VersionID=52136).
Data analysis
Data were extracted from eligible articles by author L.C.N., and all extractions were reviewed
and confirmed by at least one other researcher. The primary outcomes of interest were the
point prevalence estimates of PTSD. In addition, information on the number, age, and sex of
participants; population-level trauma exposure; sampling procedures; and the language, trans-
lation, validity, and reliability of the PTSD assessment tools were also extracted.
Pooled prevalence estimates were calculated across all studies, and then within subgroups
including by sex, assessment time frame (i.e., 1 week, 1 month, 1 year), use of a screening or
diagnostic measure, and whether populations were affected or not affected by mass-casualty
war or armed conflict. War or armed conflict was defined using the international humanitar-
ian law definition, in which armed conflict occurs between organized armed groups, govern-
mental or nongovernmental [58]. Pooled prevalences at the national and regional level were
calculated to produce a heat map of available PTSD estimates in SSA (see Fig 2). Pooled esti-
mates were calculated using the Stata version 14.2 [59] metaprop [60] command, which allows
for the inclusion of all studies, including those with 0% or 100% prevalence proportions. Meta-
prop was run using (a) a random-effects model, which assumes that differences in prevalence
estimates are not solely due to sampling error, (b) the exact confidence intervals, and (c) the
PLOS MEDICINE
Meta-analysis of prevalence of PTSD in sub-Saharan Africa
PLOS Medicine | https://doi.org/10.1371/journal.pmed.1003090 May 15, 2020 7 / 30
Freeman-Tukey double arcsine transformation to normalize the prevalence estimates prior to
pooling [60]. Q and I
2
were calculated to assess heterogeneity across all studies and within and
between subgroups [61]. Finally, a meta-regression with all a priori–defined subgroups in
which significant differences in pooled prevalence were identified was run using the Stata
metareg command [62].
Weighted prevalence estimates were used if they were reported; otherwise, raw numbers of
participants with PTSD and total sample size were included as data. If raw numbers of partici-
pants with PTSD were not reported, they were calculated from summary data reported in the
manuscripts. Prevalence estimates were calculated based on the total number of participants
included in the studies regardless of whether or not they reported experiencing a traumatic
event. In some cases, publications only reported conditional PTSD prevalence estimates (i.e.,
PTSD prevalence only among people who reported a traumatic event). In cases in which the
manuscript only reported conditional prevalence, but the overall prevalence estimate could be
calculated based on the other information provided in the manuscript, the overall prevalence
estimate was calculated and was included in the meta-analysis. Study authors were contacted
for clarification by L.C.N. in cases for which prevalence estimates were ambiguous.
Results
A total of 6,435 articles were identified through database and reference searches, of which
2,825 were unique articles (see Fig 1 for complete PRISMA flow diagram). After removing the
duplicates and 26 abstracts that could not be located, 2,799 articles remained for full review.
The coders disagreed on the eligibility of 4 articles: for 2 studies, the mean PTSD symptom
scores were reported but not estimates of diagnostic prevalence [63,64], and 2 studies met all
inclusion criteria except that they did not report PTSD scores [65,66]. After review by the first
author and in discussion with the coders, consensus was reached on eligibility. Ultimately,
2,762 articles were excluded, leaving 37 that met eligibility. The 37 articles described 25 unique
studies across 10 of the 48 SSA countries (see Table 1 and Fig 2).
Six of the studies were national surveys, including 2 from Liberia [67,68], 1 from Rwanda
[69,70], 1 from Sierra Leone [71], and 2 from South Africa [2,72–77]. The other 19 studies
were representative samples of a region within a country. Across the 25 studies, data were
Fig 2. PTSD in SSA: Pooled prevalence from countries with regionally or nationally representative data. PTSD,
posttraumatic stress disorder; SSA, sub-Saharan Africa.
https://doi.org/10.1371/journal.pmed.1003090.g002
PLOS MEDICINE
Meta-analysis of prevalence of PTSD in sub-Saharan Africa
PLOS Medicine | https://doi.org/10.1371/journal.pmed.1003090 May 15, 2020 8 / 30
Table 1. Identified studies.
Country Studies Population-
level trauma
exposure
National
or
regional
Regional,
specified
If regional, how/
why region was
selected
Year of
data
collection
Inclusion
criteria
Response
rate
Nwith
PTSD
data
PTSD measure
and scoring
Time
frame
PTSD
prevalence
(95% CI)
Women with
PTSD data
PTSD %;
sex
difference
Age: M (years)
(SD/CI); range
DRC Johnson et al.,
2010 [124]
Violence and
armed
conflict for
decades
Regional Subregions of
the North
and South
Kivu
provinces and
the Ituri
district
High levels of
violence during
the war in
Rwanda that
crossed over the
border into the
DRC
Mar 2010 Adults (18+) 98.9% 989 PSS-I in
Kiswahili;
DSM-IV
criteria
Lay
administered
1
month
50.1%
(43.8%–
56.3%)
593 60.0% Men:
44.4%
(35.4%–
53.4%)
Women:
54.0%
(46.6%–
61.5%).
p= 0.08
Mean = 40.1
(95% CI 38.2–
42.0 years)
Ethiopia Asnakew et al.,
2019 [83]
Deadly
garbage
landfill
landslide
Regional Area of the
Koshe
landslide,
Addis Ababa,
Ethiopia,
2018
Area
experienced
trauma due to
landslide
May and
June 2018
Participants
(15+ years old)
98.2% 830 PCL-C in
Amharic
1
month
37.3%
(34.1% to
40.8%)
491 59.2% Men:
31.3%
(26.4%–
36.5%)
Women:
41.5%
(37.1%–
46.0%)
p= 0.003
Mean = 33
years
(SD = 12)
Kenya Jenkins et al., 2015
[125]
Election
violence in
August 2007
and 2013
Regional The Maseno
area in
Kisumu
county,
Nyanza
Province in
western
Kenya
Endemic for
malaria. In
addition,
Nyanza
province
experienced
serious election
violence in
August 2007
6 months
in 2013
None reported 96.4% 1,146 TSQ in
English,
Kiswahili, and
Dholuo
Cutoff of 6
Restricted to
those who
experienced
trauma after
age 16
1 week 10.6%
(8.8%–
12.5%)
545 47.6% Men: 6.7%
(4.8%–
9.0%)
Women:
14.9%
(12.0%–
18.1%)
p<0.001
<30
years = 281;
30–60
years = 448;
>60
years = 171
Liberia Galea et al., 2010
[84]
Armed
violence and
conflict from
1989 to 2003
Regional Rural areas of
Nimba
County
History of civil
conflict in
Nimba County
Nov and
Dec 2008
Analyses
restricted to
those aged 19+
98.0% 1,376 HTQ in
Liberian
English
Self-reported
1 week 48.3%
(45.7%–
50.9%)
631 45.9% Not
reported
<35 y = 574
(41.7%);
>35 = 802
(58.3%)
Johnson et al.,
2008 [67]
National N/A N/A May 2008 Adults (18+),
give accurate
information
about
household
98.2% 1,661 PSS-I in
Liberian
English;
DSM-IV
criteria
Lay
administered
1
month
44.0%
(38.0%–
49.0%)
876 52.7% Men: 46%
(38%–
54%)
Women:
42%
(38%–
46%)
p= 0.17
Mean = 41
(95% CI 40–42
years)
Vinck et al., 2013
[68]
National N/A N/A Nov and
Dec 2010
Adults (18+) 93.4% 4,496 PCL-C
(language not
reported)
Cutoff of 44
Information
on PCL-C
assessors is not
provided
1
month
12.6%
(11.5%–
13.9%)
2,272 50.5% Men: 6.3%
(5.3%–
7.4%)
Women:
18.8%
(17.2%–
20.5%)
p<0.0001
Mean = 37.4
(SE = 0.26)
Nigeria Gureje et al., 2006
[98]
None
reported
Regional Yoruba-
speaking
areas: Lagos,
Ogun, Osun,
Oyo, Ondo,
Ekiti, Kogi,
and Kwara
Not reported Feb and
Nov 2002
Adults (18+),
fluent in
Yoruba
79.9% 4,984 WMH Survey
version of the
CIDI in
Yoruba
Lay
administered
12
months
0.0%
(0.0%–
0.0%)
2,552 51.2% Men: 0%
(0%–0.2%)
Women:
0% (0%–
0.1%)
p>0.95
Mean = 35
(SE = 0.37)
(Continued)
PLOS MEDICINE
Meta-analysis of prevalence of PTSD in sub-Saharan Africa
PLOS Medicine | https://doi.org/10.1371/journal.pmed.1003090 May 15, 2020 9 / 30
Table 1. (Continued)
Country Studies Population-
level trauma
exposure
National
or
regional
Regional,
specified
If regional, how/
why region was
selected
Year of
data
collection
Inclusion
criteria
Response
rate
Nwith
PTSD
data
PTSD measure
and scoring
Time
frame
PTSD
prevalence
(95% CI)
Women with
PTSD data
PTSD %;
sex
difference
Age: M (years)
(SD/CI); range
Rwanda Fodor et al., 2015
[96]
1994
genocide that
followed
years of
ethnic
tensions and
conflict
Regional Ngoma
commune in
the Huye
district in the
South
Province
Heavily affected
by the genocide
Feb and
Mar 2011
Adults (18+) 96% 500 PCL-C in
Kinyarwanda
Cutoff score of
44
Administered
by trained
native
Rwandan
college
graduates in
interview
format
1
month
21.6%
(18.1%–
25.5%)
349 75.0% Not
reported
Mean = 41.06
(14.9); 18–83
Munyandamutsa
et al., 2012 [70],
Eytan et al., 2015
[69]
National N/A N/A 3 months
in 2008
Ages 16+,
Rwandan,
fluent in
Kinyarwanda,
no mental
impairment
Not
reported
962 PTSD section
of the MINI
DSM-IV and
ICD-10 criteria
Clinician-
administered
1
month
26.1%
(23.2%–
28.9%)
567 58.9% Men:
20.5%
(16.6%–
24.8%)
Women:
30.0%
(26.2%–
33.9%)
p= 0.001
Median = 33;
range: 16–108
years
Pham et al., 2004
[126]
Regional 4 communes:
Ngoma
(known as
Butare town),
Mabanza,
Buyoga, and
Mutura)
Diversity in
region, level of
urbanization,
experience with
the genocide,
and relationship
to the ICTR
Feb 2002 Adults (18+) 99% 2,091 PCL-C
(language not
reported);
cutoff of 44
Self-reported
1
month
24.8%
(23.0%–
26.6%)
1,074 51.4% Men:
19.6%
(17.2%–
22.2%)
Women:
29.7%
(27.0%–
32.5%)
p<0.001
Mean = 36.4
years; range
18–94 years
Rugema et al.,
2015 [80]
Regional Southern
Province
Not reported Dec 2011
to Jan
2012
Aged 20–35
years,
Rwandan
99.8% 913 MINI in
Kinyarwanda
Clinician-
administered
1
month
13.6%
(11.4%–
16.0%)
477 52.0% Men: 7.1%
(4.8%–
9.9%)
Women:
19.5%
(16.0%–
23.3%)
p<0.001
20–24 = 275
(30.3%); 25–
29 = 300
(33.0%); 30–
35 = 333
(36.7%)
Sierra
Leone
Betancourt et al.,
2016 [97]
Civil war
from 1991–
2002 and
EVD
Regional Western area
rural district
and western
area urban
district
Diversity in
ethnic
composition
and degrees of
war exposure;
was epicenter of
EVD cases
during the
2014–2015
outbreak
Jan to
Apr 2015
Adults (18+) 98% 1,008 PSS-I adapted
for use in
Liberia in Krio
Lay (trained
research
assistants)
administered
1
month
6% (4%–
7%)
505 50.8% Not
reported
Mean = 34.2
years (95% CI
33.2–35.2)
Jalloh et al., 2018
[71]
Ebola National N/A N/A Jul 2015 Head of
household and
another
individual
(aged between
15 years and 24
years) or a
woman
98% 3,564 IES-6
Administered
by trained data
collectors
1 week 16%
(14.7%–
17.1%)
1,774 50% Not
reported
Median = 35
(SD = 15)
(Continued)
PLOS MEDICINE
Meta-analysis of prevalence of PTSD in sub-Saharan Africa
PLOS Medicine | https://doi.org/10.1371/journal.pmed.1003090 May 15, 2020 10 / 30
Table 1. (Continued)
Country Studies Population-
level trauma
exposure
National
or
regional
Regional,
specified
If regional, how/
why region was
selected
Year of
data
collection
Inclusion
criteria
Response
rate
Nwith
PTSD
data
PTSD measure
and scoring
Time
frame
PTSD
prevalence
(95% CI)
Women with
PTSD data
PTSD %;
sex
difference
Age: M (years)
(SD/CI); range
South
Africa
Herman et al.,
2009 [75];
Williams et al.,
2008 [127]; Atwoli
et al., 2013 [72];
Atwoli et al., 2017
[73]; Koenen et al.,
2017 [2]; Stein
et al., 2008 [74];
Myer et al., 2009
[128]; Duckers
et al., 2018 [129]
Apartheid
until 1994. In
the post-
Apartheid
era, high
rates of
nonpolitical
violence and
crime, child
abuse, IPV,
and HIV/
AIDS and TB
National N/A N/A Jan 2002
to June
2004
Adult South
Africans
residing in
households
and hostels
87.1% 4,351 CIDI version
3.0 in English,
Afrikaans,
Zulu, Xhosa,
North Sotho,
South Sotho,
and Tswana;
DSM-IV PTSD
Lay
administered
12
months
0.6%
(0.4%–
0.8%)
2,619 60.2% Men: 0.6%
(0.3%–
1.1%)
Women:
0.6%
(0.3%–
1.0%)
p= 0.92
Not reported
Machisa et al.,
2017 [85]
Regional Gauteng
Province
Not reported 2010 Adult women 79% 501 HTQ in
English, Zulu,
Sotho, and
Afrikaans
Researcher
administered
Mean
score 2.5
1 week 11.6% (9%–
14%)
495 100.0% Only
women
included
18–29 = 30%;
30–44 = 36%;
>45 = 33%
Peltzer and
Pengpid, 2019 [77]
National N/A N/A 2012 Aged 15+ 92.6% 15,201 17-item DTS
Information
on assessors is
not provided
1 week 2.1%
(1.9%–
2.3%)
8,254 54.3% Men: 2.0%
(1.7%–
2.4%)
Women:
2.3%
(2.0%–
2.6%)
p= 0.20
Mean = 36.8
(SD = 16.5)
Smit et al., 2006
[82]
Regional Periurban
settlement
outside Cape
Town, South
Africa
High HIV
prevalence
Not
reported
Aged 15+ 64.3% 645 HTQ in Xhosa
and English; a
cutoff score of
75
Self-reported
1 week 14.9%
(12.2%–
17.9%)
357 55.3% Men:
18.7%
(14.0%–
24.0%)
Women:
13.7%
(10.3%–
17.7%)
p= 0.10
Mean = 30.3
(11.9)
Topper et al., 2015
[81]
Regional Eastern Cape
Province,
urban and
semiurban
areas of
Nelson
Mandela Bay
and the
semirural
area of
Kirkwood
Not reported Not
reported
Aged 18–40
years
97.7% 977 MINI, version
6.0.0 in Xhosa,
English, and
Afrikaans;
information on
PCL-C
assessors is not
provided
1
month
10.8%
(9.0%–
13.0%)
467 47.8% Men:
11.6%
(8.9%–
14.7%)
Women:
10.1%
(7.5%–
13.2%)
p= 0.45
Not reported
(Continued)
PLOS MEDICINE
Meta-analysis of prevalence of PTSD in sub-Saharan Africa
PLOS Medicine | https://doi.org/10.1371/journal.pmed.1003090 May 15, 2020 11 / 30
Table 1. (Continued)
Country Studies Population-
level trauma
exposure
National
or
regional
Regional,
specified
If regional, how/
why region was
selected
Year of
data
collection
Inclusion
criteria
Response
rate
Nwith
PTSD
data
PTSD measure
and scoring
Time
frame
PTSD
prevalence
(95% CI)
Women with
PTSD data
PTSD %;
sex
difference
Age: M (years)
(SD/CI); range
South
Sudan
Ayazi et al., 2012
[99]; Ayazi et al.,
2014 [130]
Decades of
ongoing civil
war
Regional The Greater
Bahr el
Ghazal region
Not reported 2010 Adults (18+) 95% 1,200 HTQ with
minor
adaptations;
Arabic with
key terms in
indigenous
languages
Cutoff of 2.0
Information
on HTQ
assessors is not
provided
2
weeks
37.6%
(34.8%–
40.4%)
506 44.0% Men:
33.6%
(30.0%–
37.4%)
Women:
42.1%
(37.8%–
46.5%)
p= 0.003
18–25 = 308
(25.7%); 26–
35 = 391
(32.6%); 36–
50 = 395
(32.9%);
>50 = 89
(7.4%)
Ng et al., 2017
[131]
Regional 11 sites in 6
of South
Sudan’s 10
states:
(Central
Equatoria,
Jonglei,
Upper Nile,
Western
Equatoria,
Eastern
Equatoria,
and Lakes)
and Abyei
Diverse in terms
of ethnicity,
socioeconomic
status,
livelihood,
exposure to
conflict, and
security access
Dec 2014
to Apr
2015
Adults (18+);
South
Sudanese
99.5% 1,520 The HTQ-R in
Classical
Arabic, Juba
Arabic, Dinka,
Nuer, Shilluk,
and Bari
DSM-IV
criteria
Information
on HTQ-R
assessors is not
provided
1 week 40.7%
(38.2%–
43.2%)
773 50.9% Men:
45.2%
(41.6%–
48.9%)
Women:
36.2%
(32.8%–
39.7%)
p<0.001
Mean = 36.93
years (SD
13.90); range
18–86
Roberts et al., 2009
[132]
Regional Juba Town High numbers
of IDPs,
returned IDPs
and refugees
20–30
Nov 2007
Adults (18+) 96.2% 1,242 Adapted
version of the
HTQ in Juba
Arabic and
Bari
Mean PTSD
scores 2.0
Trained lay
administered
1 week 36.2%
(33.2%–
39.4%)
630 50.7% Men:
29.7%
(25.4%–
34.5%)
Women:
42.5%
(39.4%–
45.7%)
p<0.01
Mean = 33
years
(Continued)
PLOS MEDICINE
Meta-analysis of prevalence of PTSD in sub-Saharan Africa
PLOS Medicine | https://doi.org/10.1371/journal.pmed.1003090 May 15, 2020 12 / 30
Table 1. (Continued)
Country Studies Population-
level trauma
exposure
National
or
regional
Regional,
specified
If regional, how/
why region was
selected
Year of
data
collection
Inclusion
criteria
Response
rate
Nwith
PTSD
data
PTSD measure
and scoring
Time
frame
PTSD
prevalence
(95% CI)
Women with
PTSD data
PTSD %;
sex
difference
Age: M (years)
(SD/CI); range
Uganda Ertl et al., 2014
[79]; Neuner et al.,
2012 [78]
Twenty-one
years of war
Regional Anaka, Awer,
and Padibe
Acholi
regions
Varying degrees
of war exposure
and distance
from the largest
town, Gulu
Jul 2007
to Apr
2008
Aged 12–25 99.9% 1,113 PDS in Luo
Information
on PDS
assessors is not
provided
1
month
15% (13%–
17%)
693 62.3% Not
reported
Not reported
Mugisha et al.,
2015 [133];
Mugisha et al.,
2015 [134];
Mugisha et al.,
2016 [135]
Regional Subcounties
Lalogi and
Koro (in
Gulu district);
Amuru and
Atiak (in
Amuru
district); and
Koch Goma
and Alero (in
Nwoya
district)
Some of the
most affected
subdistricts by
the 20-year civil
war in northern
Uganda that had
a health center
2 Jan
2013 to 2
June 2013
Adults (18+),
stayed in the
area for more
than 6 months,
could fluently
speak Luo
100% 2,361 MINI in Luo
Administered
by trained
psychiatric
nurses
1
month
11.8%
(10.5%–
13.1%)
1,475 62.5% Men:
13.4%
(11.2%–
15.7%)
Women:
10.9%
(9.3%–
12.5%)
p= 0.002
18–24 = 555
(23.5%); 25–
34 = 644
(27.3%); 35–
44 = 490
(20.8%); 45–
54 = 672
(28.5%)
Pham et al., 2009
[136]
Regional 8 districts:
Amuru, Gulu,
Kitgum,
Pader, Lira,
Oyam,
Amuria, and
Soroti
Represent a
variety of ethnic
groups (Acholi,
Iteso, and
Langi) and
exposure to the
armed conflict
Mar to
June 2007
Adults (18+) 64.5% 2,867 PCL-C in 3
local languages
Cutoff score of
44
Non-clinician
administered
1
month
54% (52%–
56%)
1,417 49.5% Men: 40%
(37.5%–
42.6%)
Women:
68.3%
(65.8%–
70.7%)
p<0.001
Mean = 35.4
years
(SD = 14.35)
Vinck et al., 2007
[137]
Regional Four districts:
Gulu,
Kitgum, Lira,
and Soroti
Selected to
represent a
diversity of
ethnic
composition
(Acholi, Langi,
and Teso) and
varying degrees
of exposure to
the war
Apr to
May 2005
Adults (18+) 73% 2,389 PCL-C in
Acholi, Lango,
and Ateso
Cutoff score of
44
Information
on PCL-C
assessors is not
provided
1
month
74.3%
(72.5%–
76.0%)
1,198 50.1% Men:
64.7%
(61.9%–
67.4%)
Women:
83.3%
(81.6%–
85.8%)
p<0.001
Mean = 37
years
(SD = 13.8
years)
Abbreviations: CI, confidence interval; CIDI, Composite International Diagnostic Interview [94]; DRC, Democratic Republic of Congo; DSM-IV, Diagnostic and Statistical Manual Version 4
[138]; DTS, Davidson Trauma Scale [89]; EVD, Ebola virus disease; HTQ, Harvard Trauma Questionnaire [86]; HTQ-R, Harvard Trauma Questionnaire-Revised; ICD-10, International
Classification of Diseases, Tenth Revision [139]; ICTR, United Nations’ International Criminal Tribunal for Rwanda; IDP, internally displaced person; IES-6, Impact of Event Scale-6 [91]; IPV,
intimate partner violence; MINI, Mini International Neuropsychiatric Interview [92]; N/A, Not applicable; PCL-C, PTSD Checklist-Civilian Version [87]; PDS, Posttraumatic Stress Diagnostic
Scale [90]; PSS-I, PTSD Symptom Scale Interview [93]; PTSD, posttraumatic stress disorder; SE, standard error; TB, tuberculosis; TSQ, Trauma Screening Questionnaire [88]
https://doi.org/10.1371/journal.pmed.1003090.t001
PLOS MEDICINE
Meta-analysis of prevalence of PTSD in sub-Saharan Africa
PLOS Medicine | https://doi.org/10.1371/journal.pmed.1003090 May 15, 2020 13 / 30
available from 58,887 participants, 54% of which were women (n= 31,672 of 58,795 partici-
pants with data on sex). Some studies did not include the raw number of participants by sex,
and so the number of women was calculated based on the information available in the paper,
such as the percentage of participants who were women. Most studies enrolled any adult (men
and women aged 18 years or older). However, some studies only enrolled specific age brackets
(e.g., any person aged 12 to 25 [78,79], 20 to 35 [80], and 18 to 40 years) [81]. In addition, 3
studies enrolled any person aged 15 or older [82,83], 1 study enrolled any person 16 and older
[69,70], and 1 study enrolled any adult aged 18 or older but restricted analyses to those aged 19
and older [84]. In addition, 1 study only enrolled adult women [85]. There were no studies
that only enrolled adult men.
Across the 25 studies, presence of PTSD was assessed using translated, and in some cases
adapted, scales (see Table 1). Sixteen studies used self-reported screening instruments admin-
istered by trained lay researchers/data collectors, including the Harvard Trauma Question-
naire (HTQ) [86] (6 studies), PTSD Checklist-Civilian Version (PCL-C) [87] (6 studies),
Trauma Screening Questionnaire (TSQ) [88] (1 study), the Davidson Trauma Scale (DTS)
[89] (1 study), the Posttraumatic Stress Diagnostic Scale (PDS) [90] (1 study), and the
Impact of Event Scale-6 (IES-6) [91] (1 study). The other 9 studies used structured or semi-
structured interviews, including the Mini International Neuropsychiatric Interview (MINI)
[92] (4 studies), the PTSD Symptom Scale Interview (PSS-I) [93] (3 studies), and the Compos-
ite International Diagnostic Interview (CIDI) [94] (2 studies). The PDS was the only scale with
validated psychometrics in the population of interest [95]. The majority of studies (n= 14;
58.3%) reported on past-month prevalence. A minority of studies used a 1- or 2-week time
frame for assessment (n= 9; 36%), and only 2 studies (8%) assessed 1-year prevalence (see
Table 1).
Study quality
Overall, the methodology of the included studies had many strengths (see Table 2). Out of a
maximum total of 8, quality scores ranged from 4 (50%) to 7 (87.5%) with an average score of
6.36 (79.5%) using Loney and colleagues (1998) [56]. Sampling methods and sample sizes were
very strong. All but 2 of the 25 studies used random samples and described their sampling
methods in detail. More than 70% of the studies used an unbiased sampling frame, such as
census data, and sample sizes were large, ranging from 500 to 15,201 participants. Response
rates were very high. One study did not report a response rate, and one had a response rate of
only 64.3%, but for three-quarters of the studies response rates were >90%.
Every study used accepted clinical or research tools for measuring PTSD. While most arti-
cles acknowledged that the tools had been validated elsewhere, of note, only one study in the
meta-analysis used a tool that had been validated during the study in question or in the same
study population previously, Ertl and colleagues (2014) in Uganda [79]. The assessors and
their training to implement the instruments were well-described (all but 2 studies provided
details about the interviewers); however, reliability was inconsistently reported. Over half
(n= 14) of the studies ran and stated Cronbach’s alpha for internal reliability, while the
remaining 11 did not report it.
Overall pooled prevalence of probable PTSD
Prevalence estimates had reporting time frames of 1 week to 1 year, and estimates were highly
variable, ranging from 0% (95% CI 0%–0%) to 74% (95% CI 72%–76%), and heterogeneous
(Q = 18,326.70, df = 24, p<0.001, I
2
= 99.87%). Overall, pooled prevalence across all studies
was 22% (95% CI 13%–32%) (see Fig 2 and Fig 3). Pooled estimates were recalculated only
PLOS MEDICINE
Meta-analysis of prevalence of PTSD in sub-Saharan Africa
PLOS Medicine | https://doi.org/10.1371/journal.pmed.1003090 May 15, 2020 14 / 30
including the 20 studies that achieved at least an 80% quality score. The pooled estimate from
the high-quality studies was 20% (95% CI 11%–32%). There was no association between study
quality and prevalence (b = −0.001 [95% CI −0.11 to 0.11], p= 0.98).
Table 2. Study quality indicators.
Country Studies 1. Random
sample or
whole
population
2.
Unbiased
sampling
frame (i.e.,
census
data)
3. Adequate
sample size
(>300
participants)
4.
Measures
were the
standard
a
5.
Outcomes
measured
by unbiased
assessors
b
6.
Adequate
response
rate (70%),
refusers
described
7.
Confidence
intervals,
subgroup
analysis
8. Study
participants
described
Total
(max
8)
Percentage
of 100%
DRC Johnson et al.,
2010
1 1 1 0.5 0.5 0.5 1 1 6.5 81.3%
Ethiopia Asnakew et al.,
2019
1 0 1 0.5 1 0.5 1 1 6 75.0%
Kenya Jenkins et al.,
2015
1 1 1 0.5 0.5 0.5 1 1 6.5 81.3%
Liberia Galea et al., 2010 1 1 1 0.5 0.5 0.5 1 1 6.5 81.3%
Johnson et al.,
2008
1 1 1 0.5 0.5 0.5 1 1 6.5 81.3%
Vinck et al., 2013 1 1 1 0.5 1 0.5 1 1 7 87.5%
Nigeria Gureje et al.,
2006
1 1 1 0.5 0.5 0.5 1 1 6.5 81.3%
Rwanda Fodor et al., 2015 1 1 1 0.5 0.5 0.5 1 1 6.5 81.3%
Munyandamutsa
et al., 2012
1 0 1 0.5 0.5 0 1 1 5 62.5%
Pham et al., 2004 1 0 1 0.5 0.5 0.5 1 1 5.5 68.8%
Rugema et al.,
2015
1 1 1 0.5 0.5 0.5 1 1 6.5 81.3%
Sierra
Leone
Betancourt et al.,
2016
1 1 1 0.5 1 0.5 1 1 7 87.5%
Jalloh et al., 2018 1 1 1 0.5 1 0.5 1 1 7 87.5%
South
Africa
Herman et al.,
2009
1 1 1 0.5 0.5 0.5 1 1 6.5 81.3%
Machisa et al.,
2017
1 1 1 0.5 0.5 0.5 1 1 6.5 81.3%
Peltzer and
Pengpid, 2019
1 1 1 0.5 1 0.5 1 1 7 87.5%
Smit et al., 2006 0.5 0 1 0.5 0 0 1 1 4 50.0%
Topper et al.,
2015
1 1 1 0.5 0.5 0.5 1 1 6.5 81.3%
South
Sudan
Ayazi et al., 2012 1 1 1 0.5 1 0.5 1 1 7 87.5%
Ng et al., 2017 0 0 1 0.5 1 0.5 1 1 5 62.5%
Roberts et al.,
2009
1 1 1 0.5 1 0.5 1 1 7 87.5%
Uganda Ertl et al., 2014 1 0 1 1 1 0.5 1 1 6.5 81.3%
Mugisha et al.,
2015
1 0 1 0.5 1 1 1 1 6.5 81.3%
Pham et al., 2009 1 0.5 1 0.5 1 0.5 1 1 6.5 81.3%
Vinck et al., 2007 1 1 1 0.5 1 0.5 1 1 7 87.5%
a
PTSD measurement tools consistent with generally accepted clinical/research criteria and tool validated during study or in population previously?
b
Interviewers and their training described; internal reliability calculated.
Abbreviations: DRC, Democratic Republic of Congo; PTSD, posttraumatic stress disorder
https://doi.org/10.1371/journal.pmed.1003090.t002
PLOS MEDICINE
Meta-analysis of prevalence of PTSD in sub-Saharan Africa
PLOS Medicine | https://doi.org/10.1371/journal.pmed.1003090 May 15, 2020 15 / 30
Pooled prevalence by sex
Five studies [71,79,84,96,97] did not report prevalence estimates by sex, and so they were not
included in the analysis of prevalence by sex. Herman and colleagues (2009) [75] and Gureje
and colleagues (2006) [98] did not report prevalence by sex but did report that there was no
significant difference by sex, and so the overall mean score was included as the score for both
men and women. Machisa and colleagues (2017) [85] only sampled women, and so the data
from this study were included in the pooled analysis for women but not men.
Of the 19 studies that reported data comparing the rates of PTSD in men and women, 12
reported a significant difference (see Table 1). Of the 12 studies that reported a significant dif-
ference in the prevalence of PTSD in men compared to women, 10 reported higher rates of
PTSD in women, and 2 reported higher rates of PTSD in men. There were no significant dif-
ferences in the pooled prevalence of PTSD by sex (test of heterogeneity between groups = 0.42,
df = 1, p= 0.52). The pooled prevalence estimate for females was 25% (95% CI 14%–39%). The
pooled prevalence estimate for males was 20% (95% CI 10%–31%) (see S1 Fig). Post hoc meta-
regression indicated that the lack of a significant difference in probable PTSD prevalence by
sex persisted when predictors included conflict exposure, reporting time frame, and assess-
ment tool type (b = −0.08 [95% CI −0.19 to 0.04], p= 0.18).
Pooled prevalence by reporting time frame
The 2 WMH Survey studies [75,98] were the only ones to use a 1-year time frame. Because the
study by Ayazi and colleagues [99] was the only study that utilized a 2-week time frame, it was
grouped with those using a 1-week time frame when stratifying by assessment time frame,
since this study used the HTQ [86], which has a 1-week–assessment time frame [86]. Across
all studies, there were significant differences by assessment time frame (random heterogeneity
test between subgroups = 79.61, df = 2, p<0.001; see S2 Fig). However, this difference is
Fig 3. Overall prevalence estimates. CI, confidence interval; ES, effect size (proportion).
https://doi.org/10.1371/journal.pmed.1003090.g003
PLOS MEDICINE
Meta-analysis of prevalence of PTSD in sub-Saharan Africa
PLOS Medicine | https://doi.org/10.1371/journal.pmed.1003090 May 15, 2020 16 / 30
explained by the 2 studies that assessed PTSD over a 1-year time frame. These 2 studies (Gureje
and colleagues [98] and Herman and colleagues [75]) were part of the cross-national WMH
Surveys [100] and were the only 2 studies to use the CIDI [94]. They each had pooled preva-
lence estimates of 0% (95% CI 0%–0%). In contrast, the pooled prevalence in the past week
was 22% (95% CI 9%–38%), and the pooled prevalence in the past month was 27% (95% CI
16%–40%). There was no significant difference in the weekly versus monthly pooled preva-
lence (random heterogeneity test between subgroups = 0.28, df = 1, p0.60). The overall
pooled prevalence of probable current PTSD, defined as a period prevalence ranging from 1
week to 1 month, was 25% (95% CI 16%–36%).
Pooled prevalence by screener versus diagnostic assessment tool
There was not a significant difference in the prevalence estimates between studies that used
screening instruments (27% [95% CI 15%–41%]) compared to those using diagnostic struc-
tured and semi-structured interviews (14% [95% CI 4%–29%]; random test for heterogeneity
between subgroups = 1.76, df = 1, p= 0.18).
Pooled prevalence by exposure to war or armed conflict
There was a significant difference in pooled prevalence estimates between studies that were
conducted in regions exposed to mass-casualty war or armed conflict at any point during the
lifetime of the participants and those that were unexposed (random test for heterogeneity
between subgroups = 13.64, df = 1, p= 0.01) (see Fig 4). The pooled prevalence estimate of
studies from exposed regions was 30% (95% CI 20%–40%), while the estimate from unexposed
regions was 8% (95% CI 3%–15%). Results of the adjusted R
2
from a meta-regression indicated
that conflict exposure accounted for 22.77% of the variance in the pooled prevalence estimates.
Fig 4. Prevalence estimates by exposure to mass-casualty war or armed conflict. Random test of heterogeneity
between subgroups: 13.64, df = 1, p<0.001. CI, confidence interval; ES, effect size (proportion).
https://doi.org/10.1371/journal.pmed.1003090.g004
PLOS MEDICINE
Meta-analysis of prevalence of PTSD in sub-Saharan Africa
PLOS Medicine | https://doi.org/10.1371/journal.pmed.1003090 May 15, 2020 17 / 30
Discussion
The goal of this review was to provide a more complete picture of the population-based data
on the current prevalence of PTSD in SSA than is available from any single study. We identi-
fied 25 studies assessing 58,887 individuals in the SSA region. There were 4 key findings in the
meta-analysis: (1) The overall pooled prevalence across all studies, including 2 studies that cal-
culated a 1-year prevalence estimate, was 22% (95% CI 13%–32%). The pooled prevalence of
current symptoms (defined as duration of 1 week to 1 month) consistent with a probable diag-
nosis of PTSD in SSA was 25% (95% CI 16%–36%). (2) Prevalence estimates of current symp-
toms across the individual studies were highly variable, ranging from 2% (95% CI 2%–2%) to
74% (95% CI 72%–76%). (3) There was no significant difference in the pooled prevalence of
PTSD for men and women, for reporting time frame, or for whether the study used a screening
versus diagnostic instrument. (4) Prevalence estimates across regions differed substantially by
population-level exposure to war or armed conflict. We discuss these findings in more detail
subsequently.
The high 22% prevalence of probable PTSD found in this meta-analysis seems to be driven
in large part by the 30% prevalence of probable PTSD found across the 17 studies that were
conducted on populations exposed to war and armed conflict. The finding that populations
experiencing war or armed conflict within the lifetime of the study participants have higher
probable PTSD prevalence is consistent with expectations. Countries with higher rates of con-
flict have populations exposed to higher levels of trauma and thus more PTSD symptoms.
However, the prevalence of PTSD of 30% found in this meta-analysis still exceeds the rates
found in 3 other meta-analyses in populations exposed to war of 15% [50], 26% [101], and
24% [51]. Moreover, the 8% prevalence of PTSD in non–war-exposed populations exceeds the
4% rate of PTSD prevalence found in other population-based cross-national studies [2], sug-
gesting that there was still a substantial population-level burden of PTSD symptoms in non–
conflict-affected countries, a burden that has often been overlooked. Taken together, these
data suggest that PTSD symptoms and probable PTSD are common in SSA. The higher preva-
lence of PTSD found in this meta-analysis may be partially explained by studies finding that
people living in SSA may be disproportionately affected by individual and population-level
trauma exposure [8,10–12,14,102] and that the vast majority of people in SSA have very low
access to mental health treatment [15,16]. These factors may result in a larger effect on the
population burden of PTSD in SSA [22,23].
There was no consistent pattern in the results for sex differences in PTSD prevalence.
While 10 of the 19 studies that reported on sex differences in PTSD prevalence reported that
women had higher rates of probable PTSD, 7 studies reported no sex difference, and 2 studies
reported that men had higher rates of probable PTSD. However, there was no difference in the
overall pooled prevalence by sex. These findings suggest that sex may frequently be a critical
variable to consider when understanding PTSD prevalence but that its explanatory power may
vary by population context.
Due to large variability in prevalence estimates, there was not a significant difference in
prevalence estimates between studies that used screening instruments compared to those
using diagnostic structured and semi-structured interviews. However, the mean ratio of the
difference between studies using screening instruments (27% [95% CI 15%–41%]) and those
using diagnostic structured and semi-structured interviews (14% [95% CI 4%–29%]) was in
line with the results of a meta-regression of prevalence of PTSD in countries experiencing war
and armed conflict that found that symptom scales produced prevalence estimates that were
1.5 to 2 times higher than diagnostic tools [50]. The authors of the meta-regression suggest
that this discrepancy may occur because symptom scales do not assess clinical significance or
PLOS MEDICINE
Meta-analysis of prevalence of PTSD in sub-Saharan Africa
PLOS Medicine | https://doi.org/10.1371/journal.pmed.1003090 May 15, 2020 18 / 30
functional impairment and may therefore overestimate the prevalence of PTSD [50]. While
the lack of data on functional impairment in symptom scales is certainly problematic, studies
have found a strong correlation between elevated distress on symptom scales and functional
impairment [103–105]. Indeed, subthreshold PTSD is associated with high distress and
impairment and increased risk of suicidality globally [106]. Given that only 1 study used an
instrument for PTSD that had been validated in the population of interest, we are unable to
conclude whether symptom scales or diagnostic instruments may be under- or overestimating
PTSD.
Limitations
In producing pooled estimates, we encountered at least 4 limitations of the identified studies
that limit our ability to make interpretations from our findings. First, only 1 study (Ertl et al.,
2014 [78,79]) used a measure of PTSD symptoms whose reliability and validity had been
assessed previously in the population of interest. PTSD symptoms are highly prevalent in the
SSA countries where they have been studied. However, the lack of reliable and valid instru-
ments used in the populations studied is a major limitation for both research and practice. In a
2018 review of qualitative studies examining post-trauma symptoms across cultural contexts,
most study participants reported symptoms consistent with Western diagnostic criteria but
also reported a number of symptoms not captured in those diagnostic classifications [107],
supporting the need for local contextualization of measures. Measures that are not locally con-
textualized and valid may result in substantial under- or overreporting of trauma exposure
and PTSD symptoms. For example, a study of Zulu-speaking participants in northeastern
KwaZulu-Natal, South Africa found that the PTSD section of the Structured Clinical Interview
for Diagnostic and Statistical Manual (DSM) Disorders, Axis I, Research Version (SCID-I RV)
[108] undercounted participants who had been exposed to traumatic events by almost 20%
compared to a Zulu Culture-Specific Trauma Experience Questionnaire [109]. Pragmatic and
feasible approaches to cultural and contextual validation of measures of post-trauma symp-
toms in challenging settings have been described [110] and need to become more widely
employed prior to undertaking population surveys. Rigorous epidemiological research to
examine the predictive validity of PTSD constructs in diverse settings is also an important
priority.
Second, although data were available from a relatively large number of participants, the
studies only represent data from 10 of the 48 SSA countries, and only 6 studies provided
national-level data. There is a complete absence of national or regional population-based data
on PTSD from more than 80% of SSA countries, including those affected by ongoing fragile
and conflict-affected situations, such as Central African Republic, Mali, and Somalia. Given
the enormous heterogeneity expected not only across the continent but also within countries
and regions, this review cannot speak to rates of PTSD in any of these other regions. Thus, sub-
stantial gaps in our knowledge of PTSD prevalence in the majority of SSA remain.
These disparities in PTSD prevalence data parallel the lack of mental health data and ser-
vices broadly in SSA [1]. Despite the enormous burden that mental disorders are projected to
pose in SSA by 2050, they remain a low priority in terms of policy initiatives and research
funding [111]. Even with political will and support, many countries may have difficulty meet-
ing identified mental health needs given limited resources and competing priorities. Indeed,
although 72% of countries in the Africa region reported that they had a stand-alone mental
health policy, only 27% had allocated resources towards implementing that plan [112].
Third, as noted previously, there was high heterogeneity in the prevalence estimates with an
I
2
of more than 99%. We were therefore unable to statistically assess the risk of publication
PLOS MEDICINE
Meta-analysis of prevalence of PTSD in sub-Saharan Africa
PLOS Medicine | https://doi.org/10.1371/journal.pmed.1003090 May 15, 2020 19 / 30
bias in this meta-analysis because none of the publication bias methods provide accurate
results with more than moderate levels of heterogeneity (i.e., I
2
<50%) [113]. We are there-
fore unable to provide insight into the level of publication bias that may be present in these
results.
Finally, the goal of this review was to focus on adult PTSD, and we excluded studies that
focused exclusively or primarily on participants who are younger than 15 years old, which is
the youngest age that the World Bank considers part of the “working age population” and is
the youngest age of majority in countries throughout the world. All of the identified studies
had participants whose mean age was 18 years old or older. However, 5 studies [70,79,82,83]
included individuals who were younger than 18 years old, and therefore some of the data
come from adolescents. However, when these studies were removed from the analyses, the
results were unchanged.
In addition, there are several limitations in the methodology of our review that should be
noted. Our inclusion criteria required that studies have at least 450 participants, use a probabi-
listic sampling procedure, and report a quantitative estimate of PTSD. As a result, we excluded
smaller studies, those that focused on key populations, and those that reported only continu-
ous measures of PTSD symptoms. Our review does not, therefore, speak to subthreshold
PTSD, PTSD in specific populations (e.g. refugees, people living with HIV, students, or com-
batants), or nuances in context that may impact PTSD presentation and prevalence, such as
IDPs currently living in conflict zones compared to those living in safe environments.
Disparities between our pooled estimates and WMH surveys
The pooled prevalence of probable PTSD in this systematic review (computed from 1 week to
1 year prevalence rates) is extremely high compared to the prevalence reported by a 2017 sum-
mary of data from the WMH surveys. For example Karam and colleagues (2014) [4] reported a
WMH 12-month survey prevalence of 1.1% ranging from 3.8% in Northern Ireland to a low of
0.2–0.3% in Beijing and Shanghai in the People’s Republic of China, 0.3% in Colombia, and
0.3% in Mexico [4]. Indeed, the current PTSD pooled prevalence found in this study far
exceeds the 3.9% lifetime rate found in the WMH surveys [2]. There are at least 4 possible rea-
sons for the disparity found between the pooled PTSD prevalence estimates identified in this
meta-analysis and the prevalence estimates found in the WMH surveys. First, although our
review reports data from only 10 countries, 6 of those countries have experienced war and
armed conflict during the lifetime of the participants. Very few such countries are included in
the WMH surveys, with Iraq and Lebanon being notable exceptions [4]. However, even in Iraq
and Lebanon, the prevalence of PTSD diagnosis is very low compared with the estimates of
prevalence observed in SSA. Second, many of the studies in our pooled estimate specifically
focused on PTSD assessment as a primary aim of the study. The WMH survey, in contrast,
aims to document the population burden of all mental and behavioral disorders globally with-
out one particular focus. In addition, in most cases the survey is administered in 2 stages, and
PTSD is only included in the second stage of the survey [114]. Third, there are specific assess-
ment nuances of the CIDI [100] (WMH survey studies were the only ones that used the CIDI)
that may have impacted PTSD symptom report. First, in this review, the CIDI was the only
diagnostic interview that was lay administered. Another key issue with the CIDI, compared to
other instruments used in most of the studies, is the skip out related to reporting trauma. That
is, PTSD is only assessed in persons who reported a qualifying trauma. Thus, if a participant
does not report a trauma, either because their trauma is not on the list queried or because the
person chooses not to report such sensitive information, then these participants do not get
assessed for PTSD [114]. As a result, PTSD symptoms may go undetected. This is particularly
PLOS MEDICINE
Meta-analysis of prevalence of PTSD in sub-Saharan Africa
PLOS Medicine | https://doi.org/10.1371/journal.pmed.1003090 May 15, 2020 20 / 30
problematic when the trauma events have not been culturally and contextually validated [109].
There is some data to suggest that the results of trauma may be underreported in the WMH
surveys. For example, the lifetime exposure of sexual violence in the South African WMH Sur-
vey found a prevalence rate of rape of 2%; however, other epidemiological surveys from South
Africa have found lifetime rates of rape between 4.5% and 12% [115].
Fourth, there may be differences in the way that individuals from different countries, com-
munities, and cultures respond to items, which may result in under- or overreporting of symp-
toms. The CIDI, like many highly structured interviews, applies an algorithm of combinations
of types of symptoms following the Diagnostic and Statistical Manual version 4 (DSM-IV)
[116] in order to produce a diagnosis. The algorithms, like the diagnostic criteria in DSM-IV,
are derived predominantly from Western populations, which give weight to the core symp-
toms that are known to have salience in those settings but not necessarily in other settings.
Concerns about the cross-cultural applicability of the CIDI have been reported in Nepal [117],
Ethiopia [118], and American Indian Reservation populations [119]. Indeed, a study that used
latent-class analysis to reexamine the depression results of the WMH surveys from the US,
New Zealand, South Africa, and Nigeria found that participants in Nigeria and South Africa
who endorsed the screening questions endorsed more severe depression symptoms than par-
ticipants from the US and New Zealand who endorsed the screening questions [120]. When
these differential patterns of responding were taken into account, the prevalence estimate for
depression in Nigeria was highest (22%), whereas the results of the Nigeria WMH Survey
found that it had the lowest rate of depression (3%). Studies are needed that examine the reli-
ability and validity of the diagnostic criteria of PTSD as well as widely used assessments such
as the CIDI in SSA populations.
Conclusions and directions for future study
Methodological limitations of the extant literature as described earlier lead us to conclude that
our pooled estimate should be interpreted with caution. Given limited information on the reli-
ability and validity of the assessment tools used and the lack of data available from most coun-
tries in the region, more work is needed before strong conclusions can be made about the
population burden of PTSD in SSA. Furthermore, the identified studies provide little or no evi-
dence regarding the proportion of the population requiring specific levels of intervention and
therefore do little to inform service planning.
However, even given these limitations, our findings suggest that health systems in SSA need
to improve the identification of and access to treatment for persons with PTSD. The percent-
age of individuals with PTSD seeking treatment is globally low and is most strikingly low in
countries of low to lower-middle income and of upper-middle income [2]. Improving detec-
tion of trauma and PTSD in primary care may be an efficient strategy, given that globally, peo-
ple with PTSD are more likely to seek care in general health settings than in specialty mental
health clinics [2]. However, major efforts to scale up mental healthcare, such as the Programme
for Improving Mental Health Care (PRIME) [121]—which is focused on improving mental
healthcare in nonhumanitarian settings in several low- and middle- income countries through
integration with maternal and primary care—have not, to date, included PTSD as a target dis-
order. This is likely, in part, due to a lack of data on the prevalence of PTSD and the burden it
poses in these settings. Moreover, misconceptions about PTSD remain. For example, PTSD is
more widely recognized as a problem for refugees, in high-conflict situations, and in humani-
tarian crises. PTSD is included in the Mental Health Gap Action Programme (mhGAP) Inter-
vention Guide for humanitarian settings but not yet included in the primary mhGAP
[122,123]. However, epidemiologic data suggest that, globally, sexual violence in the context of
PLOS MEDICINE
Meta-analysis of prevalence of PTSD in sub-Saharan Africa
PLOS Medicine | https://doi.org/10.1371/journal.pmed.1003090 May 15, 2020 21 / 30
intimate partnerships is responsible for the largest burden of PTSD [5]. Thus, much founda-
tional epidemiological work remains to be done to document the burden of PTSD in SSA.
This work will be critical to inform policy, research, and treatment; above all, it will be critical
to address issues of access to care and decrease the burden of PTSD in SSA.
Supporting information
S1 PRISMA Checklist.
(DOC)
S1 Fig. Pooled prevalence by sex. Random test of herterogeneity between subgroups: 0.42,
df = 1, p= 0.052. CI, confidence interval; ES, effect size (proportion)
(TIFF)
S2 Fig. Pooled prevalence by reporting time frame. Random test of heterogeneity between
subgroups: 79.61, df = 2, p<0.001. CI, confidence interval; ES = efect size (proportion)
(TIFF)
Acknowledgments
The authors thank Emilie Wigdor, Supriya Misra, Kristen Nishimi, Fernando Martinez
Gausch, and Yuhan Cheng for their incredible assistance with identifying and coding articles,
Yuhan Cheng for helping to format this manuscript for publication, Zach Leber and Jonathan
Drummey for their guidance on creating the prevalence heat map, and OpenStreetMap con-
tributors (openstreetmap.org) for the underlying data used in the map.
Author Contributions
Conceptualization: Lauren C. Ng, Charlotte Hanlon, Soraya Seedat, Boniface Harerimana,
Bonginkosi Chiliza, Karestan C. Koenen.
Data curation: Lauren C. Ng, Anne Stevenson, Sreeja S. Kalapurakkel.
Formal analysis: Lauren C. Ng, Anne Stevenson, Sreeja S. Kalapurakkel.
Funding acquisition: Lauren C. Ng, Karestan C. Koenen.
Investigation: Lauren C. Ng, Boniface Harerimana.
Methodology: Lauren C. Ng, Anne Stevenson, Karestan C. Koenen.
Project administration: Lauren C. Ng, Karestan C. Koenen.
Resources: Lauren C. Ng, Karestan C. Koenen.
Software: Lauren C. Ng.
Supervision: Lauren C. Ng, Karestan C. Koenen.
Validation: Lauren C. Ng.
Visualization: Lauren C. Ng, Anne Stevenson.
Writing – original draft: Lauren C. Ng, Anne Stevenson, Karestan C. Koenen.
Writing – review & editing: Lauren C. Ng, Anne Stevenson, Sreeja S. Kalapurakkel, Charlotte
Hanlon, Soraya Seedat, Boniface Harerimana, Bonginkosi Chiliza, Karestan C. Koenen.
PLOS MEDICINE
Meta-analysis of prevalence of PTSD in sub-Saharan Africa
PLOS Medicine | https://doi.org/10.1371/journal.pmed.1003090 May 15, 2020 22 / 30
References
1. Whiteford HA, Degenhardt L, Rehm J, Baxter AJ, Ferrari AJ, Erskine HE, et al. Global burden of dis-
ease attributable to mental and substance use disorders: findings from the Global Burden of Disease
Study 2010. Lancet. 2013; 382(9904):1575–86. Epub 2013/09/03. https://doi.org/10.1016/S0140-
6736(13)61611-6 PMID: 23993280.
2. Koenen KC, Ratanatharathorn A, Ng L, McLaughlin KA, Bromet EJ, Stein DJ, et al. Posttraumatic
stress disorder in the World Mental Health Surveys. Psychological medicine. 2017; 47(13):2260–74.
Epub 2017/04/08. https://doi.org/10.1017/S0033291717000708 PMID: 28385165; PubMed Central
PMCID: PMC6034513.
3. Kessler RC. Posttraumatic stress disorder: the burden to the individual and to society. J Clin Psychia-
try. 2000; 61 Suppl 5:4–12; discussion 3–4. Epub 2000/04/13. PMID: 10761674.
4. Karam EG, Friedman MJ, Hill ED, Kessler RC, McLaughlin KA, Petukhova M, et al. Cumulative trau-
mas and risk thresholds: 12-month PTSD in the World Mental Health (WMH) surveys. Depression and
anxiety. 2014; 31(2):130–42. Epub 2013/08/29. https://doi.org/10.1002/da.22169 PMID: 23983056;
PubMed Central PMCID: PMC4085043.
5. Kessler RC, Aguilar-Gaxiola S, Alonso J, Benjet C, Bromet EJ, Cardoso G, et al. Trauma and PTSD in
the WHO World Mental Health Surveys. Eur J Psychotraumatol. 2017; 8(sup5):1353383. Epub 2017/
10/28. https://doi.org/10.1080/20008198.2017.1353383 PMID: 29075426; PubMed Central PMCID:
PMC5632781.
6. Global, regional, and national incidence, prevalence, and years lived with disability for 301 acute and
chronic diseases and injuries in 188 countries, 1990–2013: a systematic analysis for the Global Bur-
den of Disease Study 2013. Lancet. 2015; 15.
7. Baxter AJ, Patton G, Scott KM, Degenhardt L, Whiteford HA. Global epidemiology of mental disorders:
what are we missing? PloS one. 2013; 8(6):e65514. Epub 2013/07/05. https://doi.org/10.1371/journal.
pone.0065514 PMID: 23826081; PubMed Central PMCID: PMC3691161.
8. World Health Organization. WHO | mhGAP Mental Health Gap Action Programme [Internet]. World
Health Organization. 2008 [cited 2015 Dec 15]. Available from: http://www.who.int/mental_health/
evidence/mhGAP/en/index.html.
9. World Health Organization. Global and regional estimates of violence against women: prevalence and
health effects of intimate partner violence and non- partner sexual violence. Geneva: 2013.
10. Abrahams N, Devries K, Watts C, Pallitto C, Petzold M, Shamu S, et al. Worldwide prevalence of non-
partner sexual violence: a systematic review. Lancet. 2014; 383(9929):1648–54. Epub 2014/02/18.
https://doi.org/10.1016/S0140-6736(13)62243-6 PMID: 24529867.
11. Bundis Entwicklung Hilft. World Risk Report Analysis and prospects 2017. 2017.
12. Marc A. CONFLICT AND VIOLENCE IN THE 21ST CENTURY CURRENT TRENDS AS OBSERVED
IN EMPIRICAL RESEARCH AND STATISTICS. World Bank Group.
13. The World Bank. Fragile and conflict affected situations: World Bank Group; 2019 [cited 2019 July 11].
Available from: https://data.worldbank.org/region/fragile-and-conflict-affected-situations.
14. Fanon F. Black Skin, White Masks. New York: Grove Press; 1952.
15. Saxena S, Thornicroft G, Knapp M, Whiteford H. Resources for mental health: scarcity, inequity,and
inefficiency. Lancet. 2007; 370(9590):878–89. Epub 2007/09/07. https://doi.org/10.1016/S0140-6736
(07)61239-2 PMID: 17804062.
16. World Health Organization. Mental health atlas. Geneva: WHO Press; 2014.
17. The World Bank. Sub-Saharan Africa: World Bank Group; 2017 [cited 2018 Jan 22]. Available from:
https://data.worldbank.org/region/sub-saharan-africa.
18. The World Bank. World Bank Country and Lending Groups: The World Bank Group; 2019 [cited 2019
July 11]. Available from: https://datahelpdesk.worldbank.org/knowledgebase/articles/906519-world-
bank-country-and-lending-groups.
19. Kebede D, Alem A, Deyassa N, Shibre T, Negash A, Beyero T. Socio-demographic correlates of
depressive disorder in Butajira, rural Ethiopia. Cent Afr J Med. 2003; 49(7–8):78–83. Epub 2004/06/
25. PMID: 15214278.
20. Wang PS, Aguilar-Gaxiola S, Alonso J, Angermeyer MC, Borges G, Bromet EJ, et al. Use of mental
health services for anxiety, mood, and substance disorders in 17 countries in the WHO world mental
health surveys. Lancet. 2007; 370(9590):841–50. Epub 2007/09/11. https://doi.org/10.1016/S0140-
6736(07)61414-7 PMID: 17826169; PubMed Central PMCID: PMC2847360.
21. Eaton J, McCay L, Semrau M, Chatterjee S, Baingana F, Araya R, et al. Scale up of services for mental
health in low-income and middle-income countries. Lancet. 2011; 378(9802):1592–603. Epub 2011/
10/20. https://doi.org/10.1016/S0140-6736(11)60891-X PMID: 22008429.
PLOS MEDICINE
Meta-analysis of prevalence of PTSD in sub-Saharan Africa
PLOS Medicine | https://doi.org/10.1371/journal.pmed.1003090 May 15, 2020 23 / 30
22. Charlson FJ, Flaxman A, Ferrari AJ, Vos T, Steel Z, Whiteford HA. Post-traumatic stress disorder and
major depression in conflict-affected populations: an epidemiological model and predictor analysis.
Global Mental Health. 2016; 3:e4. https://doi.org/10.1017/gmh.2015.26 PMID: 28596873
23. Bromet EJ, Atwoli L, Kawakami N, Navarro-Mateu F, Piotrowski P, King AJ, et al. Post-traumatic
stress disorder associated with natural and human-made disasters in the World Mental Health Sur-
veys. Psychological medicine. 2017; 47(2):227–41. Epub 2016/08/31. https://doi.org/10.1017/
S0033291716002026 PMID: 27573281; PubMed Central PMCID: PMC5432967.
24. Rasmussen A, Nguyen L, Wilkinson J, Vundla S, Raghavan S, Miller KE, et al. Rates and impact of
trauma and current stressors among Darfuri refugees in Eastern Chad. The American journal of ortho-
psychiatry. 2010; 80(2):227–36. Epub 2010/06/18. https://doi.org/10.1111/j.1939-0025.2010.01026.x
PMID: 20553516; PubMed Central PMCID: PMC2920620.
25. Mels C, Derluyn I, Broekaert E, Rosseel Y. The psychological impact of forced displacement and
related risk factors on Eastern Congolese adolescents affected by war. Journal of child psychology
and psychiatry, and allied disciplines. 2010; 51(10):1096–104. Epub 2010/03/30. https://doi.org/10.
1111/j.1469-7610.2010.02241.x PMID: 20345839.
26. de Jong JT, Komproe IH, Van Ommeren M, El Masri M, Araya M, Khaled N, et al. Lifetime events and
posttraumatic stress disorder in 4 postconflict settings. JAMA: the journal of the American Medical
Association. 2001; 286(5):555–62. Epub 2001/08/14. https://doi.org/10.1001/jama.286.5.555 PMID:
11476657.
27. Salah TT, Abdelrahman A, Lien L, Eide AH, Martinez P, Hauff E. The mental health of internally dis-
placed persons: an epidemiological study of adults in two settlements in Central Sudan. The Interna-
tional journal of social psychiatry. 2013; 59(8):782–8. Epub 2012/09/18. https://doi.org/10.1177/
0020764012456810 PMID: 22982816.
28. Karunakara UK, Neuner F, Schauer M, Singh K, Hill K, Elbert T, et al. Traumatic events and symptoms
of post-traumatic stress disorder amongst Sudanese nationals, refugees and Ugandans in the West
Nile. Afr Health Sci. 2004; 4(2):83–93. Epub 2004/10/13. PMID: 15477186; PubMed Central PMCID:
PMC2141616.
29. Onyut LP, Neuner F, Ertl V, Schauer E, Odenwald M, Elbert T. Trauma, poverty and mental health
among Somali and Rwandese refugees living in an African refugee settlement—an epidemiological
study. Conflict and health. 2009; 3:6. Epub 2009/05/28. https://doi.org/10.1186/1752-1505-3-6 PMID:
19470171; PubMed Central PMCID: PMC2695430.
30. Roberts B, Ocaka KF, Browne J, Oyok T, Sondorp E. Factors associated with post-traumatic stress
disorder and depression amongst internally displaced persons in northern Uganda. BMC psychiatry.
2008; 8:38. Epub 2008/05/21. https://doi.org/10.1186/1471-244X-8-38 PMID: 18489768; PubMed
Central PMCID: PMC2397420.
31. Ndetei DM, Khasakhala L, Maru H, Pizzo M, Mutiso V, Ongecha-Owuor FA, et al. Clinical epidemiol-
ogy in patients admitted at Mathari Psychiatric Hospital, Nairobi, Kenya. Social psychiatry and psychi-
atric epidemiology. 2008; 43(9):736–42. Epub 2008/05/10. https://doi.org/10.1007/s00127-008-0360-
yPMID: 18465102.
32. Mahenge B, Likindikoki S, Stockl H, Mbwambo J. Intimate partner violence during pregnancy and
associated mental health symptoms among pregnant women in Tanzania: a cross-sectional study.
Bjog. 2013; 120(8):940–6. Epub 2013/03/08. https://doi.org/10.1111/1471-0528.12185 PMID:
23464524.
33. Whetten K, Shirey K, Pence BW, Yao J, Thielman N, Whetten R, et al. Trauma history and depression
predict incomplete adherence to antiretroviral therapies in a low income country. PloS one. 2013; 8
(10):e74771. Epub 2013/10/15. https://doi.org/10.1371/journal.pone.0074771 PMID: 24124455;
PubMed Central PMCID: PMC3790775.
34. Kinyanda E, Hoskins S, Nakku J, Nawaz S, Patel V. The prevalence and characteristics of suicidality
in HIV/AIDS as seen in an African population in Entebbe district, Uganda. BMC psychiatry. 2012;
12:63. Epub 2012/06/21. https://doi.org/10.1186/1471-244X-12-63 PMID: 22713589; PubMed Central
PMCID: PMC3395571.
35. Peltzer K, Naidoo P, Matseke G, Louw J, McHunu G, Tutshana B. Prevalence of post-traumatic stress
symptoms and associated factors in tuberculosis (TB), TB retreatment and/or TB-HIV co-infected pri-
mary public health-care patients in three districts in South Africa. Psychology, health & medicine.
2013; 18(4):387–97. Epub 2012/10/16. https://doi.org/10.1080/13548506.2012.726364 PMID:
23061988.
36. El-Sayed AM, Hadley C, Tessema F, Tegegn A, Cowan JA Jr., Galea S. Back and neck pain and psy-
chopathology in rural sub-Saharan Africa: evidence from the Gilgel Gibe Growth and Development
Study, Ethiopia. Spine (Phila Pa 1976). 2010; 35(6):684–9. Epub 2010/02/09. https://doi.org/10.1097/
BRS.0b013e3181b4926e PMID: 20139812; PubMed Central PMCID: PMC2891327.
PLOS MEDICINE
Meta-analysis of prevalence of PTSD in sub-Saharan Africa
PLOS Medicine | https://doi.org/10.1371/journal.pmed.1003090 May 15, 2020 24 / 30
37. Adewuya AO, Ologun YA, Ibigbami OS. Post-traumatic stress disorder after childbirth in Nigerian
women: prevalence and risk factors. BJOG. 2006; 113(3):284–8. Epub 2006/02/21. https://doi.org/10.
1111/j.1471-0528.2006.00861.x PMID: 16487199.
38. Chhagan MK, Mellins CA, Kauchali S, Craib MH, Taylor M, Kvalsvig JD, et al. Mental health disorders
among caregivers of preschool children in the Asenze study in KwaZulu-Natal, South Africa. Maternal
and child health journal. 2014; 18(1):191–9. Epub 2013/03/08. https://doi.org/10.1007/s10995-013-
1254-5 PMID: 23467845; PubMed Central PMCID: PMC3708993.
39. Kuo C, Operario D. Health of adults caring for orphaned children in an HIV-endemic community in
South Africa. AIDS care. 2011; 23(9):1128–35. Epub 2011/04/12. https://doi.org/10.1080/09540121.
2011.554527 PMID: 21480009; PubMed Central PMCID: PMC3139727.
40. Ertl V, Saile R, Neuner F, Catani C. Drinking to ease the burden: a cross-sectional study on trauma,
alcohol abuse and psychopathology in a post-conflict context. BMC psychiatry. 2016; 16:202. Epub
2016/06/28. https://doi.org/10.1186/s12888-016-0905-7 PMID: 27342048; PubMed Central PMCID:
PMC4921056.
41. Verelst A, De Schryver M, Broekaert E, Derluyn I. Mental health of victims of sexual violence in eastern
Congo: associations with daily stressors, stigma, and labeling. BMC women’s health. 2014; 14:106.
Epub 2014/09/10. https://doi.org/10.1186/1472-6874-14-106 PMID: 25195041; PubMed Central
PMCID: PMC4237834.
42. Mels C, Derluyn I, Broekaert E, Rosseel Y. Screening for traumatic exposure and posttraumatic stress
symptoms in adolescents in the war-affected eastern Democratic Republic of Congo. Archives of pedi-
atrics & adolescent medicine. 2009; 163(6):525–30. Epub 2009/06/03. https://doi.org/10.1001/
archpediatrics.2009.56 PMID: 19487608.
43. Othieno CJ, Okoth R, Peltzer K, Pengpid S, Malla LO. Traumatic experiences, posttraumatic stress
symptoms, depression, and health-risk behavior in relation to injury among University of Nairobi stu-
dents in Kenya. Int J Psychiatry Med. 2015; 50(3):299–316. Epub 2015/11/13. https://doi.org/10.1177/
0091217415610310 PMID: 26561275.
44. Oladeji BD, Morakinyo JJ, Gureje O. Traumatic events and post-traumatic stress symptoms among
adolescents in Ibadan. African journal of medicine and medical sciences. 2011; 40(1):23–31. Epub
2011/08/13. PMID: 21834258.
45. Neugebauer R, Fisher PW, Turner JB, Yamabe S, Sarsfield JA, Stehling-Ariza T. Post-traumatic
stress reactions among Rwandan children and adolescents in the early aftermath of genocide. Interna-
tional journal of epidemiology. 2009; 38(4):1033–45. https://doi.org/10.1093/ije/dyn375 PMID:
19204009.
46. Saile R, Ertl V, Neuner F, Catani C. Does war contribute to family violence against children? Findings
from a two-generational multi-informant study in Northern Uganda. Child abuse & neglect. 2014; 38
(1):135–46. Epub 2013/11/19. https://doi.org/10.1016/j.chiabu.2013.10.007 PMID: 24239222.
47. Okello J, Nakimuli-Mpungu E, Musisi S, Broekaert E, Derluyn I. War-related trauma exposure and mul-
tiple risk behaviors among school-going adolescents in Northern Uganda: the mediating role of
depression symptoms. Journal of affective disorders. 2013; 151(2):715–21. Epub 2013/09/10. https://
doi.org/10.1016/j.jad.2013.07.030 PMID: 24012069.
48. O’Donnell DA, Roberts WC, Schwab-Stone ME. Community violence exposure and post-traumatic
stress reactions among Gambian youth: the moderating role of positive school climate. Social psychia-
try and psychiatric epidemiology. 2011; 46(1):59–67. https://doi.org/10.1007/s00127-009-0162-x
PMID: 19921080.
49. Seedat S, Nyamai C, Njenga F, Vythilingum B, Stein DJ. Trauma exposure and post-traumatic stress
symptoms in urban African schools. Survey in CapeTown and Nairobi. The British journal of psychia-
try: the journal of mental science. 2004; 184:169–75. Epub 2004/02/03. https://doi.org/10.1192/bjp.
184.2.169 PMID: 14754831.
50. Charlson F, van Ommeren M, Flaxman A, Cornett J, Whiteford H, Saxena S. New WHO prevalence
estimates of mental disorders in conflict settings: a systematic review and meta-analysis. Lancet.
2019; 394(10194):240–8. Epub 2019/06/16. https://doi.org/10.1016/S0140-6736(19)30934-1 PMID:
31200992.
51. Hoppen TH, Morina N. The prevalence of PTSD and major depression in the global population of adult
war survivors: a meta-analytically informed estimate in absolute numbers. Eur J Psychotraumatol.
2019; 10(1):1578637. Epub 2019/03/06. https://doi.org/10.1080/20008198.2019.1578637 PMID:
30834069; PubMed Central PMCID: PMC6394282.
52. OECD (2020). Working age population (indicator). [cited 27 Apr 2020]. https://doi.org/10.1787/
d339918b-en
53. Naing L, Winn T, Rusli B. Practical issues in calculating the sample size for prevalence studies.
Archives of orofacial Sciences. 2006; 1:9–14.
PLOS MEDICINE
Meta-analysis of prevalence of PTSD in sub-Saharan Africa
PLOS Medicine | https://doi.org/10.1371/journal.pmed.1003090 May 15, 2020 25 / 30
54. Kohrt BA, Rasmussen A, Kaiser BN, Haroz EE, Maharjan SM, Mutamba BB, et al. Cultural concepts
of distress and psychiatric disorders: literature review and research recommendations for global men-
tal health epidemiology. International journal of epidemiology. 2014; 43(2):365–406. Epub 2013/12/
25. https://doi.org/10.1093/ije/dyt227 PMID: 24366490; PubMed Central PMCID: PMC3997373.
55. Moher D, Liberati A, Tetzlaff A, Altman DG, PRISMA Group. Preferred Reporting Items for Systematic
Reviews and Meta-Analyses: The PRISMA Statement. PLoS Med. 2009; 6(7):e1000097. https://doi.
org/10.1371/journal.pmed.1000097 PMID: 19621072
56. Loney PL, Chambers LW, Bennett KJ, Roberts JG, Stratford PW. Critical appraisal of the health
research literature: prevalence or incidence of a health problem. Chronic diseases in Canada. 1998;
19(4):170–6. Epub 1999/02/24. PMID: 10029513.
57. Shamliyan T, Kane RL, Dickinson S. A systematic review of tools used to assess the quality of obser-
vational studies that examine incidence or prevalence and risk factors for diseases. Journal of clinical
epidemiology. 2010; 63(10):1061–70. Epub 2010/08/24. https://doi.org/10.1016/j.jclinepi.2010.04.014
PMID: 20728045.
58. International Committee of the Red Cross (ICRC). How is the Term "Armed Conflict" Defined in Inter-
national Humanitarian Law? March 2008. [cited 27 Apr 2020]. Available from: https://www.icrc.org/en/
doc/assets/files/other/opinion-paper-armed-conflict.pdf.
59. StataCorp. Stata Statistical Software: Release 14. College Station, TX: StataCorp LP.; 2015.
60. Nyaga VN, Arbyn M, Aerts M. Metaprop: a Stata command to perform meta-analysis of binomial data.
Archives of public health = Archives belges de sante publique. 2014; 72(1):39. Epub 2014/01/01.
https://doi.org/10.1186/2049-3258-72-39 PMID: 25810908; PubMed Central PMCID: PMC4373114.
61. Higgins JPT, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ:
British Medical Journal. 2003; 327(7414):557–60. https://doi.org/10.1136/bmj.327.7414.557 PMID:
12958120.
62. Harbord RM, Higgins JPT. Meta-regression in Stata. Stata Journal. 2008; 8(4):493–519.
63. Pence BW, Shirey K, Whetten K, Agala B, Itemba D, Adams J, et al. Prevalence of psychological
trauma and association with current health and functioning in a sample of HIV-infected and HIV-unin-
fected Tanzanian adults. PloS one. 2012; 7(5):e36304. Epub 2012/05/19. https://doi.org/10.1371/
journal.pone.0036304 PMID: 22606252; PubMed Central PMCID: PMC3351441.
64. Kohli A, Perrin N, Mpanano RM, Case J, Murhula CM, Binkurhorhwa AK, et al. Socialinteraction in the
aftermath of conflict-related trauma experiences among women in Walungu Territory, Democratic
Republic of Congo. Global public health. 2015; 10(1):55–70. Epub 2014/11/05. https://doi.org/10.
1080/17441692.2014.972426 PMID: 25363537; PubMed Central PMCID: PMC4262667.
65. Ayazi T, Lien L, Eide AH, Jenkins R, Albino RA, Hauff E. Disability associated with exposure to trau-
matic events: results from a cross-sectional community survey in South Sudan. BMC public health.
2013; 13:469. Epub 2013/05/16. https://doi.org/10.1186/1471-2458-13-469 PMID: 23672785;
PubMed Central PMCID: PMC3658891.
66. Jenkins R, Njenga F, Okonji M, Kigamwa P, Baraza M, Ayuyo J, et al. Psychotic symptoms in Kenya—
prevalence, risk factors, and relationship with common mental disorders. Int J Environ Res Public
Health. 2012; 9(5):1748–56. Epub 2012/07/04. https://doi.org/10.3390/ijerph9051748 PMID:
22754470; PubMed Central PMCID: PMC3386585.
67. Johnson K, Asher J, Rosborough S, Raja A, Panjabi R, Beadling C, et al. Association of combatant
status and sexual violence with health and mental health outcomes in postconflict Liberia. JAMA: the
journal of the American Medical Association. 2008; 300(6):676–90. https://doi.org/10.1001/jama.300.
6.676 PMID: 18698066.
68. Vinck P, Pham PN. Association of exposure to intimate-partner physical violence and potentially trau-
matic war-related events with mental health in Liberia. Social science & medicine. 2013; 77:41–9.
Epub 2012/12/12. https://doi.org/10.1016/j.socscimed.2012.10.026 PMID: 23219850.
69. Eytan A, Munyandamutsa N, Nkubamugisha PM, Gex-Fabry M. Long-term mental health outcome in
post-conflict settings: Similarities and differences between Kosovo and Rwanda. The International
journal of social psychiatry. 2015; 61(4):363–72. Epub 2014/08/26. https://doi.org/10.1177/
0020764014547062 PMID: 25145869.
70. Munyandamutsa N, Nkubamugisha PM, Gex-Fabry M, Eytan A. Mental and physical health in Rwanda
14 years after the genocide. Social psychiatry and psychiatric epidemiology. 2012; 47(11):1753–61.
https://doi.org/10.1007/s00127-012-0494-9 PMID: 22402589
71. Jalloh MF, Li W, Bunnell RE, Ethier KA, O’Leary A, Hageman KM, et al. Impact of Ebola experiences
and risk perceptions on mental health in Sierra Leone, July 2015. BMJ global health. 2018; 3(2):
e000471. Epub 2018/04/03. https://doi.org/10.1136/bmjgh-2017-000471 PMID: 29607096; PubMed
Central PMCID: PMC5873549.
PLOS MEDICINE
Meta-analysis of prevalence of PTSD in sub-Saharan Africa
PLOS Medicine | https://doi.org/10.1371/journal.pmed.1003090 May 15, 2020 26 / 30
72. Atwoli L, Stein DJ, Williams DR, McLaughlin KA, Petukhova M, Kessler RC, et al. Trauma and post-
traumatic stress disorder in South Africa: analysis from the South African Stress and Health Study.
BMC psychiatry. 2013; 13:182. Epub 2013/07/04. https://doi.org/10.1186/1471-244X-13-182 PMID:
23819543; PubMed Central PMCID: PMC3716970.
73. Atwoli L, Stein DJ, King A, Petukhova M, Aguilar-Gaxiola S, Alonso J, et al. Posttraumatic stress disor-
der associated with unexpected death of a loved one: Cross-national findings from the world mental
health surveys. Depression and anxiety. 2017; 34(4):315–26. Epub 2016/12/07. https://doi.org/10.
1002/da.22579 PMID: 27921352; PubMed Central PMCID: PMC5661943.
74. Stein DJ, Seedat S, Herman A, Moomal H, Heeringa SG, Kessler RC, et al. Lifetime prevalence of psy-
chiatric disorders in South Africa. The British journal of psychiatry: the journal of mental science. 2008;
192(2):112–7. Epub 2008/02/05. https://doi.org/10.1192/bjp.bp.106.029280 PMID: 18245026;
PubMed Central PMCID: PMC2718689.
75. Herman AA, Stein DJ, Seedat S, Heeringa SG, Moomal H, Williams DR. The South African Stress and
Health (SASH) study: 12-month and lifetime prevalence of common mental disorders. South African
medical journal = Suid-Afrikaanse tydskrif vir geneeskunde. 2009; 99(5 Pt 2):339–44. Epub 2009/07/
11. PMID: 19588796; PubMed Central PMCID: PMC3191537.
76. Williams DR, Gonzalez HM, Williams S, Mohammed SA, Moomal H, Stein DJ. Perceived discrimina-
tion, race and health in South Africa. Social science & medicine. 2008; 67(3):441–52. https://doi.org/
10.1016/j.socscimed.2008.03.021 PMID: 18486292; PubMed Central PMCID: PMC2643465.
77. Peltzer K, Pengpid S. High physical activity is associated with post-traumatic stress disorder among
individuals aged 15 years and older in South Africa. The South African journal of psychiatry: SAJP: the
journal of the Society of Psychiatrists of South Africa. 2019; 25:1329. Epub 2019/11/21. https://doi.org/
10.4102/sajpsychiatry.v25i0.1329 PMID: 31745437; PubMed Central PMCID: PMC6851996.
78. Neuner F, Pfeiffer A, Schauer-Kaiser E, Odenwald M, Elbert T, Ertl V. Haunted by ghosts: prevalence,
predictors and outcomes of spirit possession experiences among former child soldiers and war-
affected civilians in Northern Uganda. Social science & medicine. 2012; 75(3):548–54. Epub 2012/05/
15. https://doi.org/10.1016/j.socscimed.2012.03.028 PMID: 22580073.
79. Ertl V, Pfeiffer A, Schauer-Kaiser E, Elbert T, Neuner F. The challenge of living on: psychopathology
and its mediating influence on the readjustment of former child soldiers. PloS one. 2014; 9(7):
e102786. Epub 2014/07/24. https://doi.org/10.1371/journal.pone.0102786 PMID: 25054324; PubMed
Central PMCID: PMC4108375.
80. Rugema L, Mogren I, Ntaganira J, Krantz G. Traumatic episodes and mental health effects in young
men and women in Rwanda, 17 years after the genocide. BMJ open. 2015; 5(6):e006778. Epub 2015/
06/26. https://doi.org/10.1136/bmjopen-2014-006778 PMID: 26109109; PubMed Central PMCID:
PMC4480039.
81. Topper K, van Rooyen K, Grobler C, van Rooyen D, Andersson LM. Posttraumatic Stress Disorder
and Barriers to Care in Eastern Cape Province, South Africa. Journal of traumatic stress. 2015; 28
(4):375–9. Epub 2015/08/14. https://doi.org/10.1002/jts.22032 PMID: 26271019.
82. Smit J, Myer L, Middelkoop K, Seedat S, Wood R, Bekker LG, et al. Mental health and sexual risk
behaviours in a South African township: a community-based cross-sectional study. Public health.
2006; 120(6):534–42. Epub 2006/05/11. https://doi.org/10.1016/j.puhe.2006.01.009 PMID:
16684549.
83. Asnakew S, Shumet S, Ginbare W, Legas G, Haile K. Prevalence of post-traumatic stress disorder
and associated factors among Koshe landslide survivors, Addis Ababa, Ethiopia: a community-based,
cross-sectional study. BMJ open. 2019; 9(6):e028550. https://doi.org/10.1136/bmjopen-2018-028550
PMID: 31256034
84. Galea S, Rockers PC, Saydee G, Macauley R, Varpilah ST, Kruk ME. Persistent psychopathology in
the wake of civil war: long-term posttraumatic stress disorder in Nimba County, Liberia. American jour-
nal of public health. 2010; 100(9):1745–51. Epub 2010/07/17. https://doi.org/10.2105/AJPH.2009.
179697 PMID: 20634461; PubMed Central PMCID: PMC2920997.
85. Machisa MT, Christofides N, Jewkes R. Mental ill health in structural pathways to women’s experi-
ences of intimate partner violence. PloS one. 2017; 12(4):e0175240. Epub 2017/04/07. https://doi.org/
10.1371/journal.pone.0175240 PMID: 28384241; PubMed Central PMCID: PMC5383260.
86. Mollica R, Caspi-Yavin Y, Lavelle J. The Harvard Trauma Questionnaire (HTQ) manual: Cambodian,
Laotian, and Vietnamese Versions. Torture Quarterly Journal on Rehabilitation of Torture Victims and
Prevention of Torture. 1996;(Supplement 1):19–42.
87. Weathers F, Litz B, Herman D, Huska J, Keane T. The PTSD checklist-civilian version (PCL-C). Bos-
ton, MA: National Center for PTSD. 1994.
PLOS MEDICINE
Meta-analysis of prevalence of PTSD in sub-Saharan Africa
PLOS Medicine | https://doi.org/10.1371/journal.pmed.1003090 May 15, 2020 27 / 30
88. Brewin CR, Rose S, Andrews B, Green J, Tata P, McEvedy C, et al. Brief screening instrument for
post-traumatic stress disorder. The British journal of psychiatry: the journal of mental science. 2002;
181:158–62. Epub 2002/08/02. https://doi.org/10.1017/s0007125000161896 PMID: 12151288.
89. Davidson JR, Book SW, Colket JT, Tupler LA, Roth S, David D, et al. Assessment of a new self-rating
scale for post-traumatic stress disorder. Psychological medicine. 1997; 27(1):153–60. Epub 1997/01/
01. https://doi.org/10.1017/s0033291796004229 PMID: 9122295.
90. Foa EB, Cashman L, Jaycox LH, Perry K. The validation of a self-report measure of posttraumatic
stress disorder: The Posttraumatic Diagnostic Scale. Psychological assessment. 1997; 9(4):445–51.
http://dx.doi.org/10.1037/1040-3590.9.4.445.
91. Thoresen S, Tambs K, Hussain A, Heir T, Johansen VA, Bisson JI. Brief measure of posttraumatic
stress reactions: impact of Event Scale-6. Social psychiatry and psychiatric epidemiology. 2010; 45
(3):405–12. Epub 2009/05/30. https://doi.org/10.1007/s00127-009-0073-x PMID: 19479171.
92. Sheehan DV, Lecrubier Y, Sheehan KH, Amorim P, Janavs J, Weiller E, et al. The Mini-International
Neuropsychiatric Interview (M.I.N.I.): the development and validation of a structured diagnostic psychi-
atric interview for DSM-IV and ICD-10. J Clin Psychiatry. 1998; 59 Suppl 20:22–33;quiz 4–57. Epub
1999/01/09. PMID: 9881538.
93. Foa EB, Riggs DS, Dancu CV, Rothbaum BO. Reliability and validity of a brief instrument for assessing
post-traumatic stress disorder. 1993. 1994-15951-001.
94. Robins LN, Wing J, Wittchen HU, Helzer JE, Babor TF, Burke J, et al. The Composite International
Diagnostic Interview. An epidemiologic Instrument suitable for use in conjunction with different diag-
nostic systems and in different cultures. Archives of general psychiatry. 1988; 45(12):1069–77. Epub
1988/12/01. https://doi.org/10.1001/archpsyc.1988.01800360017003 PMID: 2848472.
95. Ertl V, Pfeiffer A, Saile R, Schauer E, Elbert T, Neuner F. Validation of a mental health assessment in
an African conflict population. Psychological assessment. 2010; 22(2):318–24. Epub 2010/06/10.
https://doi.org/10.1037/a0018810 PMID: 20528059.
96. Fodor KE, Pozen J, Ntaganira J, Sezibera V, Neugebauer R. The factor structure of posttraumatic
stress disorder symptoms among Rwandans exposed to the 1994 genocide: A confirmatory factor
analytic study using the PCL-C. J Anxiety Disord. 2015; 32:8–16. Epub 2015/04/04. https://doi.org/10.
1016/j.janxdis.2015.03.001 PMID: 25840139.
97. Betancourt TS, Brennan RT, Vinck P, VanderWeele TJ, Spencer-Walters D, Jeong J, et al. Associa-
tions between Mental Health and Ebola-Related Health Behaviors: A Regionally Representative
Cross-sectional Survey in Post-conflict Sierra Leone. PloS medicine. 2016; 13(8):e1002073. Epub
2016/08/10. https://doi.org/10.1371/journal.pmed.1002073 PMID: 27505186; PubMed Central
PMCID: PMC4978463.
98. Gureje O, Lasebikan VO, Kola L, Makanjuola VA. Lifetime and 12-month prevalence of mental disor-
ders in the Nigerian Survey of Mental Health and Well-Being. The British journal of psychiatry: the jour-
nal of mental science. 2006; 188:465–71. Epub 2006/05/02. https://doi.org/10.1192/bjp.188.5.465
PMID: 16648534.
99. Ayazi T, Lien L, Eide AH, Ruom MM, Hauff E. What are the risk factors for the comorbidity of posttrau-
matic stress disorder and depression in a war-affected population? A cross-sectional community study
in South Sudan. BMC psychiatry. 2012; 12:175. Epub 2012/10/23. https://doi.org/10.1186/1471-244X-
12-175 PMID: 23083301; PubMed Central PMCID: PMC3534332.
100. Kessler RC, Ustun TB. The World Mental Health (WMH) Survey Initiative Version of the World Health
Organization (WHO) Composite International Diagnostic Interview (CIDI). Int J Methods Psychiatr
Res. 2004; 13(2):93–121. Epub 2004/08/07. https://doi.org/10.1002/mpr.168 PMID: 15297906.
101. Morina N, Stam K, Pollet TV, Priebe S. Prevalence of depression and posttraumatic stress disorder in
adult civilian survivors of war who stay in war-afflicted regions. A systematic review and meta-analysis
of epidemiological studies. Journal of affective disorders. 2018; 239:328–38. Epub 2018/07/22. https://
doi.org/10.1016/j.jad.2018.07.027 PMID: 30031252.
102. World Health Organization. mhGAP in Ethiopia: Proof of Concept. 2013. [cited 27 Apr 2020]. Available
from: https://www.who.int/mental_health/mhgap/mhgap_ethiopia_proof_of_concept_2013.pdf.
103. Bolton P, Wilk CM, Ndogoni L. Assessment of depression prevalence in rural Uganda using symptom
and function criteria. Social psychiatry and psychiatric epidemiology. 2004; 39(6):442–7. Epub 2004/
06/19. https://doi.org/10.1007/s00127-004-0763-3 PMID: 15205728.
104. Bolton P, Neugebauer R, Ndogoni L. Prevalence of depression in rural Rwanda based on symptom
and functional criteria. The Journal of nervous and mental disease. 2002; 190(9):631–7. Epub 2002/
10/03. https://doi.org/10.1097/00005053-200209000-00009 PMID: 12357098.
105. Mollica RF, McInnes K, Sarajlic N, Lavelle J, Sarajlic I, Massagli MP. Disability associated with psychi-
atric comorbidity and health status in Bosnian refugees living in Croatia. JAMA: the journal of the
PLOS MEDICINE
Meta-analysis of prevalence of PTSD in sub-Saharan Africa
PLOS Medicine | https://doi.org/10.1371/journal.pmed.1003090 May 15, 2020 28 / 30
American Medical Association. 1999; 282(5):433–9. Epub 1999/08/12. https://doi.org/10.1001/jama.
282.5.433 PMID: 10442658.
106. McLaughlin KA, Koenen KC, Friedman MJ, Ruscio AM, Karam EG, Shahly V, et al. Subthreshold post-
traumatic stress disorder in the world health organization world mental health surveys. Biological psy-
chiatry. 2015; 77(4):375–84. Epub 2014/05/21. https://doi.org/10.1016/j.biopsych.2014.03.028 PMID:
24842116; PubMed Central PMCID: PMC4194258.
107. Michalopoulos LM, Meinhart M, Yung J, Barton SM, Wang X, Chakrabarti U, et al. Global Posttrauma
Symptoms: A Systematic Review of Qualitative Literature. Trauma, violence & abuse.
2018:1524838018772293. Epub 2018/04/28. https://doi.org/10.1177/1524838018772293 PMID:
29699456.
108. First M, B., Spitzer R, L., Gibbon M, Williams Janet BW. Structured Clinical Interview for DSM-IV-TR
Axis I Disorders, Research Version, Non-patient Edition (SCID-I/NP). New York, NY: New York State
Psychiatric Institute, 2002.
109. Madigoe T, Burns J, Zhang M, Subramaney U. Towards a culturally appropriate trauma assessment
in a South African Zulu community. Psychol Trauma. 2017; 9(3):274–81. Epub 2017/05/02. https://doi.
org/10.1037/tra0000231 PMID: 28459270.
110. Haroz EE, Bass JK, Lee C, Murray LK, Robinson C, Bolton P. Adaptation and testing of psychosocial
assessment instruments for cross-cultural use: an example from the Thailand Burma border. BMC
psychology. 2014; 2(1):31. Epub 2015/02/17. https://doi.org/10.1186/s40359-014-0031-6 PMID:
25685351; PubMed Central PMCID: PMC4317135.
111. Hyman SE. The Unconscionable Gap Between What We Know and What We Do. Science Transla-
tional Medicine. 2014; 6(253):253cm9. https://doi.org/10.1126/scitranslmed.3010312 PMID:
25210058
112. Mental health atlas 2017. Geneva: World Health Organization; 2018. Licence: CC BY-NC-SA 3.0 IGO.
113. van Aert RCM, Wicherts JM, van Assen MALM. Publication bias examined in meta-analyses from psy-
chology and medicine: A meta-meta-analysis. PloS one. 2019; 14(4):e0215052–e. https://doi.org/10.
1371/journal.pone.0215052 PMID: 30978228.
114. The World Health Organization. Download the WHO WMH-CIDI Instruments 2017 [cited 2019 March
7]. Available from: https://www.hcp.med.harvard.edu/wmhcidi/download-the-who-wmh-cidi-
instruments/.
115. Jewkes R, Abrahams N. The epidemiology of rape and sexual coercion in South Africa: an overview.
Social science & medicine. 2002; 55(7):1231–44. Epub 2002/10/09. https://doi.org/10.1016/s0277-
9536(01)00242-8 PMID: 12365533.
116. American Psychiatric Association. Diagnostic and statistical manual of mental disorders (DSM-IV).
4th ed. Washington DC: American Psychiatric Association; 2000.
117. van Ommeren M, Sharma B, Makaju R, Thapa S, de Jong J. Limited Cultural Validity of the Composite
International Diagnostic Interview’s Probe Flow Chart. Transcult Psychiatry. 2000; 37(1):119–29.
https://doi.org/10.1177/136346150003700107.
118. Gelaye B, Williams MA, Lemma S, Deyessa N, Bahretibeb Y, Shibre T, et al. Diagnostic validity of the
composite international diagnostic interview (CIDI) depression module in an East African population.
Int J Psychiatry Med. 2013; 46(4):387–405. Epub 2013/01/01. https://doi.org/10.2190/PM.46.4.e
PMID: 24922989; PubMed Central PMCID: PMC4058648.
119. Beals J, Manson SM, Whitesell NR, Mitchell CM, Novins DK, Simpson S, et al. Prevalence of major
depressive episode in two American Indian reservation populations: unexpected findings with a struc-
tured interview. The American journal of psychiatry. 2005; 162(9):1713–22. Epub 2005/09/02. https://
doi.org/10.1176/appi.ajp.162.9.1713 PMID: 16135632.
120. Scorza P, Masyn K, Salomon JA, Betancourt TS. The impact of measurement differences on cross-
country depression prevalence estimates: A latent transition analysis. PloS one. 2018; 13(6):
e0198429. Epub 2018/06/08. https://doi.org/10.1371/journal.pone.0198429 PMID: 29879167;
PubMed Central PMCID: PMC5991686.
121. Lund C, Tomlinson M, De Silva M, Fekadu A, Shidhaye R, Jordans M, et al. PRIME: a programme to
reduce the treatment gap for mental disorders in five low- and middle-income countries. PLoS medi-
cine. 2012; 9(12):e1001359. Epub 2013/01/10. https://doi.org/10.1371/journal.pmed.1001359 PMID:
23300387; PubMed Central PMCID: PMC3531506.
122. World Health Organization. mhGAP intervention guide for mental, neurological and substance use dis-
orders in non-specialized health settings: Mental Health Gap Action Programme (mhGAP). Geneva:
World Health Organization; 2010.
123. World Health Organization and United Nations High Commissioner for Refugees. mhGAP Humanitar-
ian Intervention Guide (mhGAP-HIG): clinical management of mental, neurological and substance use
conditions in humanitarian emergencies (Version 1.0). Geneva: 2015.
PLOS MEDICINE
Meta-analysis of prevalence of PTSD in sub-Saharan Africa
PLOS Medicine | https://doi.org/10.1371/journal.pmed.1003090 May 15, 2020 29 / 30
124. Johnson K, Scott J, Rughita B, Kisielewski M, Asher J, Ong R, et al. Association of sexual violence
and human rights violations with physical and mental health in territories of the Eastern Democratic
Republic of the Congo. JAMA: the journal of the American Medical Association. 2010; 304(5):553–62.
https://doi.org/10.1001/jama.2010.1086 PMID: 20682935.
125. Jenkins R, Othieno C, Omollo R, Ongeri L, Sifuna P, Mboroki JK, et al. Probable Post Traumatic
Stress Disorder in Kenya and Its Associated Risk Factors: A Cross-Sectional Household Survey. Int J
Environ Res Public Health. 2015; 12(10):13494–509. Epub 2015/10/31. https://doi.org/10.3390/
ijerph121013494 PMID: 26516877; PubMed Central PMCID: PMC4627045.
126. Pham PN, Weinstein HM, Longman T. Trauma and PTSD symptoms in Rwanda: implications for atti-
tudes toward justice and reconciliation. JAMA: the journal of the American Medical Association. 2004;
292(5):602–12. https://doi.org/10.1001/jama.292.5.602 PMID: 15292086.
127. Williams DR, Herman A, Stein DJ, Heeringa SG, Jackson PB, Moomal H, et al. Twelve-month mental
disorders in South Africa: prevalence, service use and demographic correlates in the population-
based South African Stress and Health Study. Psychological medicine. 2008; 38(2):211–20. Epub
2007/10/02. https://doi.org/10.1017/S0033291707001420 PMID: 17903333; PubMed Central PMCID:
PMC2718686.
128. Myer L, Stein DJ, Grimsrud AT, Herman A, Seedat S, Moomal H, et al. DSM-IV-defined common men-
tal disorders: association with HIV testing, HIV-related fears, perceived risk and preventive behaviours
among South African adults. South African medical journal = Suid-Afrikaanse tydskrif vir genees-
kunde. 2009; 99(5 Pt 2):396–402. PMID: 19588804.
129. Du¨ckers MLA, Alisic E, Brewin CR. A vulnerability paradox in the cross-national prevalence of post-
traumatic stress disorder. British Journal of Psychiatry. 2018; 209(4):300–5. Epub 01/02. https://doi.
org/10.1192/bjp.bp.115.176628
130. Ayazi T, Lien L, Eide A, Swartz L, Hauff E. Association between exposure to traumatic events and anx-
iety disorders in a post-conflict setting: a cross-sectional community study in South Sudan. BMC psy-
chiatry. 2014; 14:6. Epub 2014/01/15. https://doi.org/10.1186/1471-244X-14-6 PMID: 24410951;
PubMed Central PMCID: PMC3893536.
131. Ng LC, Lopez B, Pritchard M, Deng D. Posttraumatic stress disorder, trauma, and reconciliation in
South Sudan. Social psychiatry and psychiatric epidemiology. 2017; 52(6):705–14. Epub 2017/04/13.
https://doi.org/10.1007/s00127-017-1376-y PMID: 28401274; PubMed Central PMCID:
PMC5510537.
132. Roberts B, Damundu EY, Lomoro O, Sondorp E. Post-conflict mental health needs: a cross-sectional
survey of trauma, depression and associated factors in Juba, Southern Sudan. BMC psychiatry. 2009;
9:7. Epub 2009/03/06. https://doi.org/10.1186/1471-244X-9-7 PMID: 19261192; PubMed Central
PMCID: PMC2656503.
133. Mugisha J, Muyinda H, Wandiembe P, Kinyanda E. Prevalence and factors associated with Posttrau-
matic Stress Disorder seven years after the conflict in three districts in northern Uganda (The Wayo-
Nero Study). BMC psychiatry. 2015; 15:170. Epub 2015/07/25. https://doi.org/10.1186/s12888-015-
0551-5 PMID: 26205192; PubMed Central PMCID: PMC4513792.
134. Mugisha J, Muyinda H, Malamba S, Kinyanda E. Major depressive disorder seven years after the con-
flict in northern Uganda: burden, risk factors and impact on outcomes (The Wayo-Nero Study). BMC
psychiatry. 2015; 15:48. Epub 2015/04/17. https://doi.org/10.1186/s12888-015-0423-z PMID:
25881310; PubMed Central PMCID: PMC4376071.
135. Mugisha J, Muyinda H, Kagee A, Wandiembe P, Mpugu SK, Vancampfort D, et al. Prevalence of sui-
cidal ideation and attempt: associations with psychiatric disorders and HIV/AIDS in post-conflict North-
ern Uganda. Afr Health Sci. 2016; 16(4):1027–35. Epub 2017/05/10. https://doi.org/10.4314/ahs.
v16i4.20 PMID: 28479896; PubMed Central PMCID: PMC5398450.
136. Pham PN, Vinck P, Stover E. Returning home: forced conscription, reintegration, and mental health
status of former abductees of the Lord’s Resistance Army in northern Uganda. BMC psychiatry. 2009;
9:23. Epub 2009/05/19. https://doi.org/10.1186/1471-244X-9-23 PMID: 19445689; PubMed Central
PMCID: PMC2693494.
137. Vinck P, Pham PN, Stover E, Weinstein HM. Exposure to war crimes and implications for peace build-
ing in northern Uganda. JAMA: the journal of the American Medical Association. 2007; 298(5):543–54.
https://doi.org/10.1001/jama.298.5.543 PMID: 17666675.
138. American Psychiatric Association. Diagnostic and statistical manual of mental disorders (DSM).
Washington, DC: American Psychiatric Association. 1994.
139. World Health Organization. ICD-10: international statistical classification of diseases and related
health problems: tenth revision. 2nd ed: World Health Organization; 2004.
PLOS MEDICINE
Meta-analysis of prevalence of PTSD in sub-Saharan Africa
PLOS Medicine | https://doi.org/10.1371/journal.pmed.1003090 May 15, 2020 30 / 30
